Oilfield Review

Summer 2005

Wellbore Cleanout
Rapid Petrophysical Answers
Fibers for Hydraulic Fracturing
Casing Directional Drilling
 ve at:
rs ck

y
/o ba
ur
om ed
.c fe
lb ur
ld yo
ie s
ilf e u

.s
w giv
w ase
.o
le
w
P

Previous survey results Please tell us what you

Topics readers want Responses think about Oilfield Review
New technology applications
Logging, formation evaluation
Drilling
Completions
The Oilfield Review communicates technical
Stimulation
advances in the oil and gas industry to interested
Satisfied with topics
professionals. Your participation in our reader
Production/reservoir engineering
survey will help us make the journal more relevant,
Market overviews, trends
informative and accessible. The on-line survey
Integrated case studies
takes just a few minutes, and your responses will
Seismic data be kept confidential.
Exploration
Fluids and testing
Simulation
Cementing
0 20 40 60

Topics readers do not want Responses

Satisfied with topics

Nontechnical topics
Overly technical topics
Sales/ads/propaganda
Equations
Topics available elsewhere
Wireline topics
Biographies
No case study
Successes only
Drilling
Geophysics
Geology
0 25 50 75

Please give us your feedback at: www.oilfield.slb.com/orsurvey

05_OR_003_0
Cultivating the Technical Talent of the Future
In a business that has a reputation for being ruthlessly
cyclic, it is relatively comforting to know that we are cur-
rently in an up cycle of great activity and likely to remain
there for some years. The worlds oil and gas production is
strained by demand, and oil prices continue to rise. Ask
most people in this business and they will tell you they have
never been busier.
So far, so good. But looking to the future, we see looming
capacity shortages not just in technology and equipment,
but also in that most precious resource of allpeople. For
years Schlumberger has based its vision on three key
principles: people, technology and profitabilityI suspect
most exploration and production businesses claim similar
guidelines. It is not surprising that the category people
heads the list, because no business succeeds without talent.
Most worrisome for some, however, is the availability of
technical talent.
There has been much talk in recent years about the
impending crew change, which refers to the fact that a
majority of the industrys technical talent is approaching
retirement while too few young experts are being trained
to replace them. Exacerbated by todays strong growth, the
effects of the crew change are beginning to be felt, but this
talent shortage phenomenon is primarily seen in the West.
The number of petroleum engineering graduates entering
the workplace from US universities dropped dramatically
in the mid- to late 1980s and never recovered, but the
number graduating in Asia-Pacific countries, to mention
just one area, is now huge. There is no crew change in
those emerging economies.
At Schlumberger, we see two essential components to
securing tomorrows technical talent. Firstand this was
a bet we made 25 years agois to hire engineers and sci-
entists in every country where we worked, in numbers
roughly proportional to our activity in each country. Then
we opted to treat everyone the same, from training, to
career, to compensation. That is why today we are a veritable
United Nations of engineers, and we do not worry about a
crew change of our own.
The second component is deeper and grew naturally
from the first. It comprises a portfolio of educational
initiatives in all countries where we worksome relating
to our business, others addressing social needs. It started
with simple university relations to back up our recruiting,
but this activity now encompasses 140 engineering schools
worldwide. We promote numerous sabbatical programs for
professors, liaisons between African scientists and top
Western universities, a program called Faculty for the
Future that funds women professors, and more programs
too numerous to mention here. Contributing to university
excellence remains a high priority.
Recognizing that we could contribute at all stages in the
educational chain, we founded the volunteer-based
Schlumberger Excellence in Educational Development
(SEED) program that connects under-resourced schools
to the Internet and provides them with a comprehensive,
multilingual science education Web site and access to
practicing scientists and engineers. It has taken time, but
we have now connected 131 schools in 32 countries. In
Africa, 30,000 schoolchildren have access to this program.
In several countries, we have brought many children together
in workshops to study environmental and other issues.
The Schlumberger Foundation is another mechanism at
our disposal to promote education, particularly science
education. This foundation supports students from develop-
ing countries, helping them attend United World Colleges,
a respected worldwide educational network, and also
maintains the traveling Lab-in-the-Lorry program that
brings experiments and scientists to children.
The technical talent of tomorrow is everywhere to be
seen. Just give everyone the chance to become tomorrows
engineer and scientistwe believe that is the secret to
securing technical talent for the future.
Chakib Sbiti
Executive Vice President, Schlumberger Oilfield Services
Paris, France
Chakib Sbiti, Executive Vice President of Schlumberger Oilfield Services (OFS),
manages oilfield technology development and all operations in the OFS business
segment worldwide. Prior to assuming this position in 2003, he was president,
Middle East and Asia, Schlumberger OFS for two years. He also served as
director of personnel, Oilfield Services, Paris; and vice president, Wireline &
Testing, Europe, Africa and the Mediterranean. Chakib joined Schlumberger in
1981 as a field engineer after studying electrical engineering in France.
1
 Schlumberger

Oilfield Review
Executive Editor 4 Integrated Wellbore Cleanout Systems: Improving
Mark A. Andersen Efficiency and Reducing Risk
Advisory Editor
Gretchen M. Gillis
Accumulation of sand and other materials in a wellbore may
have catastrophic effects on oil flow. By integrating process,
Senior Editors chemistry and downhole tool technology, engineers now safely
Mark E. Teel
Matt Garber and efficiently remove this debris from wells. Case histories
from North America, the North Sea and Malaysia demonstrate
Editors how careful planning and process integration save time,
Don Williamson
Roopa Gir reduce cost and risk, and help operators return wells to
Matt Varhaug production more quickly.
Contributing Editors
Rana Rottenberg
Joan Mead
Erik B. Nelson
Julian Singer 14 Spectroscopy: The Key to Rapid, Reliable
Petrophysical Answers
Design/Production
Herring Design A new log interpretation technique obtains rock properties
Steve Freeman
from elemental concentrations with nearly automatic
Illustration processing of data from modern spectroscopy and conventional
Tom McNeff
Mike Messinger
logging tools. The method provides fast, objective and reliable
George Stewart interpretations to make completion decisions. Case studies
from Egypt, Venezuela and the North Sea illustrate the technique.
Printing
Wetmore Printing Company
Curtis Weeks

On the cover:

Accumulation of sand and solids in

wellbores significantly impairs oil and
gas production. A field specialist lowers
a coiled tubing-conveyed jetting tool
into a wellbore during a field test at the
Schlumberger Reservoir Completions
Technology Center, Rosharon, Texas, Useful links: Address editorial Address distribution inquiries to:
USA. The new tool design significantly correspondence to: Matt Garber
improves solids removal efficiency Schlumberger Oilfield Review Schlumberger Cambridge Research
during wellbore-cleanout operations. www.slb.com 1325 S. Dairy Ashford High Cross, Madingley Road
Houston, Texas 77077 USA Cambridge, England CB3 0EL
Oilfield Review Archive (1) 281-285-7847 (44) 1223 325 377
www.slb.com/oilfieldreview Fax: (1) 281-285-1537 Fax: (44) 1223 361 473
E-mail: editorOilfieldReview@slb.com E-mail: mgarber@cambridge.oilfield.slb.com
Oilfield Glossary
www.glossary.oilfield.slb.com

2
 Summer 2005
Volume 17
Number 2

34 New Fibers for Hydraulic Fracturing Advisory Panel

Syed A. Ali
An innovative fracturing fluid incorporates special fibers to Chevron Energy Technology Co.
improve proppant transport. This new technology significantly Houston, Texas, USA
reduces the fluid viscosity required for effective proppant trans- Abdulla I. Al-Kubaisy
port. This article reviews fiber-assisted transport and presents Saudi Aramco
Ras Tanura, Saudi Arabia
case histories that illustrate how fibers help engineers design
George King
optimal fracturing treatments and improve stimulation results. BP
Houston, Texas
Eteng A. Salam
PERTAMINA
Jakarta, Indonesia
Y.B. Sinha
New Delhi, India
Sjur Talstad
44 Using Casing to Drill Directional Wells Statoil
Stavanger, Norway
During the past five years, more than 350 vertical intervals
and about a dozen inclined sections have been drilled with Richard Woodhouse
Independent consultant
casing. Among oil and gas companies, however, there is Surrey, England
interest in applying this technique to drill in problematic
offshore fields where high-angle wells are common. We review
the use of larger tubulars for drilling operations and present
results from work in south Texas, including a multiwell
comparison of casing directional drilling with both downhole
motors and rotary steerable systems.

62 Contributors

65 New Books

66 Coming in Oilfield Review

Address editorial
correspondence to:
Oilfield Review
1325 S. Dairy Ashford
Houston, Texas 77077 USA
(1) 281-285-7847
Fax: (1) 281-285-1537
E-mail: editorOilfieldReview@slb.com
Address distribution inquiries to:
Matt Garber
Schlumberger Cambridge Research
High Cross, Madingley Road
Cambridge, England CB3 0EL
(44) 1223 325 377
Fax: (44) 1223 361 473
E-mail:
mgarber@cambridge.oilfield.slb.com
Oilfield Review subscriptions
are available from:
Oilfield Review Services
Barbour Square, High Street
Tattenhall, Chester CH3 9RF England
(44) 1829-770569
Fax: (44) 1829-771354
E-mail: orservices@t-e-s.co.uk
www.oilfieldreview.com
Annual subscriptions, including postage,
are 180.00 US dollars, subject to
exchange-rate fluctuations.
Oilfield Review is published quarterly by
Schlumberger to communicate technical
advances in finding and producing hydro-
3
 Integrated Wellbore Cleanout Systems:
Improving Efciency and Reducing Risk

Accumulation of sand and solids in wellbores signicantly impairs oil and gas
production. In fact, nearly half of all coiled tubing operations involve well
cleanouts to remove debris. Innovative integration of hardware, software, uid
cleanout systems and treatment monitoring helps engineers reduce the cost and
risk of wellbore cleanout operations and return wells to production more quickly.

Azhar Ali Movement of sand and accumulation of debris
PETRONAS Carigali Sdn Bhd (PCSB) can have a considerable impact on uid ow. On
Kerteh, Malaysia the surface, a river can deposit so much silt that
it blocks its own ow, changing its course and
Curtis G. Blount perhaps threatening farmland and communities.
ConocoPhillips
Similarly, downhole in a well, inux of sand can
Anchorage, Alaska, USA
impair or stop the ow of oil from a reservoir.
Stephen Hill Sand ll and debris are not new wellbore
Jai Pokhriyal problems. Generations of oileld engineers have
Xiaowei Weng faced the challenge of keeping wellbores cleaned
Sugar Land, Texas, USA out. In 1901, the Jennings Oil Company Clement
No. 1 well in southwestern Louisiana, USA,
M. J. Loveland gushed oil at an estimated rate of 7,000 bbl/d
ConocoPhillips [1,113 m3/d].1 Unfortunately for these early oil
Kuparuk, Alaska pioneers, prosperity was short-lived. After seven
hours of production, formation sand plugged
Shahril Mokhtar
more than 1,000 ft [305 m] of casing, stiing oil
Kemaman, Malaysia
production along with dreams of riches and
Jessica Pedota wealth. Attempts to remove the sand from this
Prudhoe Bay, Alaska wellbore eventually failed and the prospect
was abandoned.
Mads Rdsj Around the same time, oil prospectors in
BP Norge AS Texas began using a novel technique to deal with
Stavanger, Norway oil-production declinethe torpedo.2 A torpedo
man carefully lowered substantial quantities of
Radovan Rolovic nitroglycerine down the wellbore. Once the
Stonehouse, England nitroglycerine containers reached their target, a
weight was dropped in the hole, initiating a
Wei Zhou
sequence of events that climaxed in a
Stavanger, Norway
spectacular explosion, and with any luck,
For help in preparation of this article, thanks to Marc Allcorn, stimulated the well, removing debris from the
Sugar Land, Texas, USA; and Markus Andre Karlsen, wellbore and reinitiating the ow of oil.
Bergen, Norway.
CoilCADE and PowerCLEAN are marks of Schlumberger.
Today, engineers use safer and more efcient
methods to remove sand and other debris from a
wellbore. In this article, case histories from
North America, the North Sea and Malaysia
demonstrate how carefully designed and
integrated wellbore cleanout processes save
time, reduce cost and risk, and improve
operational efficiency, while also allowing
operators to produce more oil.
Moving Solids up the Wellbore
Wellbore ll is a major concern for operators
throughout the world. This production-inhibiting
problem is commonly dealt with through coiled
tubing (CT) interventions. However, as wellbores
and completions become more complex and as
reserves are produced under increasingly difcult
conditions, there are environments where
conventional coiled tubing cleanout techniques
are not adequate for effective ll removal.
Wellbore cleanouts were among the first
applications for coiled tubing services. Global
estimates suggest that nearly 50% of CT jobs
are performed to remove mobile solids and
debris, such as produced sand or residual
proppant from hydraulic-fracturing treatments.3
Continued developments in CT conveyance
systems generally have allowed operators to keep
pace with increasingly greater well depths, more
tortuous boreholes and more difcult downhole
environmental conditions.4

4 Oileld Review
 The most common technique for deviated
wellbore cleanout uses a jetting tool conveyed
downhole by CT. While pumping cleanout uid
down the tubing, the tool is lowered, or washed,
into the sand or other debris, often called ll. At
some distance, or bite, into the ll, downward
motion is stopped. While continuing to circulate
cleanout uid, the jetting tool is slowly pulled
Summer 2005
uphole some distance in a process known as
sweeping. How large a bite is taken and how far
the tool is pulled uphole is dependent on many
parameters including ow rate, the type of ll,
tubing and casing sizes, the cleanout uid used,
nozzle design, bottomhole pressure and wellbore
trajectory. Occasionally, the sweep will have to be
brought all the way back to surface before taking
the next bite. Once the ll has been swept
upward to a predetermined depth, the tool is
1. Adapted from an article by Shelia Esthay, Jennings Daily
News, http://www.dnr.state.la.us/cons/rst-well.ssi
(accessed March 2, 2005).
2. Olien RM: The Oil Field Shooters, http://www.
texancultures.utsa.edu/hiddenhistory/pages1/
OilenOilShooter.htm (accessed March 2, 2005).
3. Rolovic R, Weng X, Hill S, Robinson G, Zemlak K and
Najafov J: An Integrated System Approach to Wellbore
Cleanouts with Coiled Tubing, paper SPE 89333,
presented at the SPE/ICoTA Coiled Tubing Conference
and Exhibition, The Woodlands, Texas, USA,
March 2324, 2004.
4. For more on coiled tubing: Afghoul AC, Amaravadi S,
Boumali A, Calmeto JCN, Lima J, Lovell J, Tinkham S,
Zemlak K and Staal T: Coiled Tubing: The Next Generation,
Oileld Review 16, no. 1 (Spring 2004): 3857.
5
 A returned to bottom, taking the next bite of ll. The
process is repeated until all ll has been mobilized
and removed from the wellbore (left).
The jetting tool, or wash nozzle, is generally
designed to create uid turbulence that helps
mobilize and suspend solid particles. However,
for inclined wellbores, turbulence decreases as
distance from the nozzles increases, and solids
often form beds on the lower side of a wellbore as
they fall, or slip, from suspension. As the height
of this solids bed increases, less of the wellbore
cross section is available for ow, so uid velocity
across the surface of the bed increases until it
reaches a critical mobilization velocity. Once this
velocity is achieved, all or a portion of the ll
B disperses, is remixed with the cleanout uid and
is transported toward the surface, often forming
a new bed farther up the hole.
As the jetting tool moves upward toward a
newly formed bed, turbulence generated by the
jetting action also helps to mobilize the ll,
transporting it uphole until solids again settle.
The cycle repeats, pushing the bed uphole as the
CT is pulled up the wellbore. If the CT speed is
too fast or the jetting nozzle is inappropriate for
the application, solids will be bypassed and
unevenly distributed along the wellbore,
resulting in only a partial cleanout and the need
for further remedial treatment. This problem
may also occur when ow rates are too low or the
C
carrier uid is incorrectly designed.

D
< Steps in the cleanout process. A typical well-
bore cleanout process involves several steps.
First, coiled tubing conveys the cleanout tool to
the top of the ll (A). In image B, the tool enters
the ll while circulating, is washing and mobilizing
the solids, and has taken a bite. Then, in image C,
a preplanned bite length has been reached and
the jetting tool is being pulled up toward the liner
top, initiating the sweeping process. In image D,
the ll is being swept through a portion of the
critical angle (40 to 65 degrees) section of hole.
Generally, once solids are swept to the top of the
liner, the nozzle is returned to bottom, the next
bite is taken, and the process repeats until all
solids have been removed from the wellbore.

6 Oileld Review
Integrating Cleanout Systems
Engineers consider many factors when designing
wellbore cleanouts, including well-completion
geometry, wellbore deviation, cleanout fluid
properties, uid ow rate, circulating pressure
limits, bottomhole pressure and temperature,
the type of solids that must be removed, and the
length along which solids must be transported.
Most often, higher ow rates, smaller completion
sizes, lighter and more angularly shaped solids,
lower deviations and downhole temperatures,
and shorter distances for solids transport lead to
easier cleanouts. However, at angles between 40
and 65 degrees, the effects of well inclination can
make almost any wellbore difcult to clean.5
Schlumberger began integrating wellbore
cleanout systems in 2002 at the Schlumberger
Integrated Productivity and Conveyance Center
(IPC) in Sugar Land, Texas. Engineers rst used
ow-loop data to validate and improve earlier
theoretical models and computer algorithms
(right). Realizing that no single aspect of the
cleanout process determines success or failure,
engineers exploited system synergies and
developed the integrated PowerCLEAN
engineered fill removal system. Software
applications, cleanout fluids, jetting-tool and
nozzle design and solids-removal monitoring
were combined into one system enabling
engineers to design cost-effective cleanout
solutions for sand, bauxite and other debris
under a wide range of wellbore conditions,
including wells with large casing sizes, high
> Large-scale ow-loop simulations. Engineers at IPC used the 7.0-in. transparent ow loop (top)
temperatures and difcult borehole trajectories.
and various coiled tubing sizes to evaluate solids transport for various uids and wash nozzle
The basic techniques behind modern CT congurations at deviations ranging from 45 to 75 from vertical. Cleanout efciency was evaluated
wellbore cleanout operations are all similar. True while varying the type of ll, annular velocities and solids loading. Tests also helped optimize nozzle
differentiation exists in the integration of key design for maximum penetration rate, particle suspension and sweeping speeds (bottom).
technical elements such as software, cleanout
uids, nozzles and solids monitoring.
SoftwareThe PowerCLEAN job-design
software serves as the integrating platform for By adjusting operating procedures, engineers difcult wellbore environments. To address this
wellbore cleanout optimization. For any given set ensure that the solids bed height will not exceed critical need, Schlumberger researchers
of wellbore and operating conditions, the a predetermined portion of the wellbore cross- developed the PowerCLEAN uid system.
software evaluates and optimizes cleanout uids sectional area, thereby minimizing friction and Engineers carefully considered the impli-
with respect to a series of variables, including tubing drag, equivalent circulating density cations of thermal effects on viscosity and
the maximum uid ow rate for a maximum (ECD), and the risk of stuck tubing.6 subsequent hole-cleaning efciency. Although
allowable circulating pressure; bottomhole Cleanout uidsFluids used in wellbore velocity plays a more important role in transport
pressure limitations; maximum CT run-in-hole cleanout operations were often developed for efciency under dynamic conditions, increasing
(RIH) speed and bite length when penetrating other oilfield operations, such as hydraulic fluid viscosity can forestall static settling.7
the fill; solids bed formation and behavior fracturing and gravel packing. In CT operations, Higher uid viscosities tend to increase frictional
relative to sweeping requirements; optimal CT cleanout performance demands on uid systems pressures and reduce ow rates at the expense of
pulling speed for sweeping; and sweep length are high. Hydraulic diameters are often small and
5. Rolovic et al, reference 3.
before taking the next bite of ll. require that engineers balance solids-transport
6. Equivalent circulating density is the effective density
Additional parameters may be set in the efficiency requirements and fluid viscosity exerted by a circulating uid against the formation that
design software to ensure a safe, efcient and against ow rates, and bottomhole temperatures takes into account the pressure drop in the annulus
above the point being considered.
problem-free cleanout. For example, the and pressures. These and other demands make 7. Rolovic et al, reference 3.
software can predict the height of solids beds many existing cleanout fluids inadequate in
that form on the low side of an inclined wellbore.

Summer 2005 7
 0.5-gal/bbl
Rheology of Various Cleanout Fluids PowerCLEAN gel Solids Transport Length
90 1.75-lbm/bbl welan (20/40 Sand at 30 ft/min Fluid Velocity)
Viscosity at 170 sec-1, cP

80 1.75-lbm/bbl xanthan
70 1.75-lbm/bbl guar
60
50

Fluid
40
30
20
10
0 0.00 0.25 0.50 0.75 1.00 1.25
50 100 150 200 250 300 350 Fluid Friction Pressure in 1.5-in. OD
Normalized transport length
Straight Coiled Tubing
Temperature, F
500
0.25-gal/bbl

Pressure drop per 1,000 ft, psl

450
> Evaluating cleanout uids. Laboratory analysis PowerCLEAN gel
400
350 1.75-lbm/bbl xanthan
shows that the PowerCLEAN uid exhibits
300 1.75-lbm/bbl guar
thermal stability to just below 325F (orange
curve - left). Laboratory tests have shown that 250
200
circulating friction pressures of PowerCLEAN 150
gel (orange) are low when compared with those Water
100 1.75-lbm/bbl welan
of common cleanout uids (middle). In this test, 50
a low-friction solution of water and friction 1.75-lbm/bbl guar
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 1.75-lbm/bbl xanthan
reducer is shown for comparison purposes (light
Flow rate, bbl/min 0.5-gal/bbl PowerCLEAN gel
blue curve). Also, when compared with xanthan-
1.05-gal/bbl friction reducer
base uid (pink), the PowerCLEAN uid (orange)
shows a 100% improvement in carrying capacity
at lower concentrations (right).

Circulating pressure
Corrected depth
Pump rate
Total volume of solids removed
> Monitoring solids removal from the Solids return rate
wellbore. The solids monitoring system PowerCLEAN Real-Time Data Output
uses acoustic signals to monitor the 800 6,000 2.0
amount of solids being removed from 4,000
the wellbore. The measuring device is
noninvasive and attaches to the return 5,000
line from the wellhead (top left and
Total volume solids removed, lbm

1.5
right). A computer interface monitors
Solids return rate, lbm/min

700 3,000
the device throughout the job. Data 4,000
Pump rate, bbl/min

output (right) shows the solids return

Pressure, psi
Depth, ft

rate versus time (red) and an estimate

3,000 1.0
of total solids removed (black). Unusual 2,000
changes in the data alert engineers to 600
potential problems during the job. 2,000
0.5 1,000
1,000
500

0 0 0
13:07:19 15:53:59 18:40:39 21:27:19 00:13:59
Time

8 Oileld Review
effective hole cleaning. Thermal effects can have
signicant deleterious effects on polymer uids,
reducing viscosity and limiting static suspension
capacity (previous page, top).
Engineers at IPC used horizontal ow loops to
investigate particle-settling velocity under
various owing conditions. The viscosity prole of Performance of Various Nozzles
a solution containing PowerCLEAN gel exhibited (Water, 60-degree Deviation)

CT pulling out of hole speed, ft/min

50
pronounced shear thinning. Further tests showed
that the fluid system provides acceptable 40
viscosity at temperatures up to 325F [163C].
30
In some cases advanced uid systems are not
needed and ordinary uids, such as water, guar, 20
hydroxyethylcellulose (HEC), xanthan, welan 10
and viscoelastic surfactant-base uids, can be
used effectively with the PowerCLEAN system. 0
20 40 60 80 100
An important factor in this process is selection of Annular fluid velocity, ft/min
the correct fluid for a given application,
PowerCLEAN nozzle
complementing velocity requirements, nozzle Forward and backward jets
design and wellbore conditions. Forward jets only
NozzlesAvailable designs include nozzles Backward jets only
that jet both forward and backward, those that
jet forward only, those that jet backward only and
those that can be switched on demand from
forward to backward. Any of these combinations
may include radial swirl-inducing features. IPC
engineers designed new nozzles using
theoretical studies and empirical cleanout tests
in 3.5- and 7-in. test loops. The nozzles are
designed to ensure complete and efficient
removal of solids from most wellbore
congurations using uids ranging from water to
viscosied cleanout uids.
PowerCLEAN nozzles have no moving parts
and provide continuous jetting to create a
> Washing ll from the wellbore. The PowerCLEAN nozzle (bottom right) outperforms other nozzle
swirling effect. Nozzle focus, direction, size and
designs. In laboratory tests using a 7.5-in. [190.5-mm] ow loop, higher achievable pumping rates and
spacing are specically designed for wellbore annular velocities coupled with the swirling effects (left) achieved by the nozzle design help keep
cleanouts of unconsolidated fill, optimizing solids in suspension longer, allowing the CT to be pulled at faster rates, saving time and improving
available fluid energy for particle lift and efciency (top right).
suspension (right). The pressure drop across the
PowerCLEAN nozzle is relatively small, typically
100 to 400 psi [689 to 2,758 kPa] at ow rates
between 1 and 3 bbl/min [159 and 477 L/min].
The small pressure drop across the nozzle allows
for higher ow rates and uid velocities in the sensor location as a function of time. Observing Wells operated by ConocoPhillips in the
wellbore, which are essential for effective the trend of solids returning to surface during a Kuparuk River Unit on the North Slope of Alaska,
removal of wellbore ll. cleanout job provides a means of verifying the USA, often have wellbore fill that hampers
Solids monitoringEnsuring that solids are performance of PowerCLEAN systems. Potential production and increases operating cost at some
being removed from the wellbore at predicted problems can be anticipated and corrective point in their life cycle.8 Wellbore trajectories
rates is critical to job success. An important action taken. can be tortuous; in some cases, undulations more
component of the PowerCLEAN system is the
8. Loveland MJ and Pedota J: Case History: Efcient
solids monitoring device, an acoustic sensor that Cleaning Undulating Trajectories in Alaska Coiled-Tubing Sand Cleanout in a High-Angle Well Using
measures the energy associated with the Integrating wellbore cleanout system components a Complete Integrated Cleaning System, paper SPE
94179, presented at the SPE/ICoTA Coiled Tubing
collisions of solids on the inside surface of a pipe allows engineers to successfully remove solids Conference and Exhibition, The Woodlands, Texas,
(previous page, bottom). This energy is processed and debris from wellbores that were previously April 1213, 2005.
to detect the volume of solids passing by the considered too complex for cleanout or those in
which remedial treatments were not considered
cost-effective.

Summer 2005 9
than 140 ft [43 m] from crest to trough make
sand removal efforts difcult (right).
Early in 2003, drillers completed a well across
a 5,000-ft [1,524-m] horizontal section of the low-
pressure West Sak sand. With the assistance of a
jet pump, the well initially produced up to
660 bbl/d [105 m3/d] of oil.
In September 2003, the well was shut in to
change the artificial lift system. During the
workover, the slickline encountered ll near the
top of the liner at 6,580 ft [2,006 m]. During the
next month, Schlumberger eld specialists ran
CT into the wellbore, tagging ll at 8,775 ft
[2,675 m] coiled tubing measured depth (CTMD).
Although slick-water with biopolymer gel pills
and slick-diesel combined with gelled diesel pills
were pumped through the CT to remove wellbore
debris, no significant amount of solids was
brought to surface.9 Later, a review of the
running-weight log indicated that the CT had not
tagged sand, but had reached its sliding-friction
limit, or a condition called helical lockup.
In November 2003, the cleanout attempt was
repeated with larger outside diameter (OD) CT.
The CT eld specialist encountered greater than
normal resistance while reentering the well,
which indicated that sand was distributed along
the length of the wellbore. A solid sand plug was
tagged just above the liner top at 6,521 ft
[1,987 m] CTMD. A diesel-base cleanout uid
was pumped down the tubing at 2.1 bbl/min
[333.8 L/min] while taking 100-ft [30.5-m] bites
into the ll before sweeping to the bottom of the
production tubing, or tubing tail. At 7,449 ft
[2,270 m], returns were lost and the CT was
immediately pulled out of the wellbore.
While pulling the CT to the surface, the eld
specialist noted high CT overpull, indicating that
some solids were being left behind along the
wellbore and were sliding down the tubing.
However, as the jetting tool approached surface,
returns were regained and engineers observed a
signicant amount of sand, wetted with gelled
diesel, being returned to surface. Following this
cleanout, the well produced for about a month
before sanding off again.
Engineers from ConocoPhillips and
Schlumberger planned a third cleanout, this
time using the PowerCLEAN integrated cleanout
system. The wellbore cleanout design modules
from CoilCADE coiled tubing design and
evaluation software allowed engineers to
evaluate several locally available cleanout uids,
including 2% potassium chloride [KCl], welan-
base, xanthan-base, diesel, gelled diesel and the
PowerCLEAN gel system. Because of low
bottomhole pressures (BHP), all uid options
10
0
1,000
True vertical depth, ft
2,000
3,000
8,750 ft 11,250 ft
4,000
4 12-in. tubing string
5,000
6,000
2,000 0 -2,000 -4,000
Horizontal displacement, ft
Packer/liner hanger/
sealbore assembly
5 12-in. liner, blank
2,300 ft
140 ft
9 58-in. casing shoe at 6,767 ft 5 12-in. liner, perforated
at every other joint
> A difcult wellbore trajectory in Alaska. The highly stratied nature of the multiple-target sands in
one of the West Sak sands of the Kuparuk River Unit, Alaska, led ConocoPhillips to drill Well IC-172
with a twisted and highly undulating borehole (top right). At 6,521 ft measured depth (3,930 feet
[1,198 m] true vertical depth) the well opens up to a 512-in. preperforated liner that runs to 11,970 ft
[3,648 m]. The undulating nature of the wellbore allows for signicant solids-bed buildup and makes
cleaning difcult.
required gas lift, either from natural gas, nozzle. The model also indicated that a single-
nitrogen or both. Due to the undulating geometry sweep operation was possible with a penetration
of this well, the exact concentration of fill rate of 7.2 ft/min [2.2 m/min] and bites of 124 ft
was unknown. [37.8 m]. Each bite would need to be circulated
For uid volume comparison, engineers for 14 minutes before taking the next.
assumed cleanouts in 500-ft [152-m] increments, During execution, an unexpected increase in
starting at 6,521 ft measured depth. Single-sweep wellhead pressure occurred. Rather than risk
cleanout simulations predicted that use of the losing returns, ConocoPhillips and Schlumberger
PowerCLEAN gel would allow completion of engineers reevaluated the job design and reduced
cleanout operations in 6 hours using 1,000 bbl the ow rate to 3 bbl/min. The remodeled design
[159 m3] of uid and 220,000 ft3 [6,230 m3] of based on this new ow rate decreased nitrogen
nitrogen. Xanthan gels would require about ow to 800 ft3/min [22.6 m3/min], slowed the
24 hours, 5,200 bbl [826 m3] of uid and 740,000 ft3 penetration rate to 7 ft/min [2.1 m/min] and
[20,956 m3] of nitrogen, while welan uids would reduced the bite size to 120 ft [36.6 m]. Reducing
need 29 hours, 5,200 bbl of uid and as much as the ow rate precluded a single-sweep circulating
920,000 ft3 [26,054 m3] of nitrogen. As for diesel- cleanout, so engineers reverted to a multiple-
base uids, the high time estimates for a single- sweep process, bringing each sweep to the
sweep cleanout and uid volume requirements liner top.
precluded further consideration. Throughout the job, the Schlumberger eld
Before the production liner was reached, gas engineer monitored the solids-removal rate using
hydrate and multiple sand bridges were cleaned the solids monitoring system, verifying the
from the production tubing.10 The PowerCLEAN efciency of the cleanout design and solids-
software model recommended a uid ow rate carrying capacity of the PowerCLEAN system.
of 4.6 bbl/min [731 L/min] with 900 ft3/min Unlike the earlier conventional cleanout systems,
[25.5 m3/min] of nitrogen through the optimized no heavy sand loads were seen while jetting out
the last 1,000 ft to surface. Periodic manual
Oileld Review
sampling of uid returns veried the accuracy of
the automated solids monitoring system.
A total of about 3,000 bbl [477 m3] of
PowerCLEAN fluid, 11,120 gal [42 m3] of
nitrogen followed by about 500 bbl [79 m3] of
diesel were pumped. The surface uid-handling
equipment had limited capacity for nitrogen
removal, so the PowerCLEAN uid could not be
recirculated and a higher than expected uid
volume was required. Improvements in degassing
methods are expected to signicantly reduce
PowerCLEAN fluid volume requirements on
future jobs. After the cleanout, CT running forces
predicted by CoilCADE modeling closely
matched actual measured values, indicating that
no sand ll remained across the cleaned liner.
Experience gained working with ConocoPhillips
in Alaska helped Schlumberger engineers ne-
tune the PowerCLEAN software modules to more
accurately simulate and plan the entire wellbore
cleanout process. Initial post-job production
UNITED KINGDOM
> Proppant removal in the North Sea. In the Valhall eld, centered approximately between Norway, Denmark, Germany and the UK
in the North Sea (bottom right), engineers at BP spend about a third of their time (top) dealing with post-stimulation wellbore cleanup.
Summer 2005
rates from this well were around 1,000 bbl/d,
later stabilizing at 500 bbl/d [79 m3/d] of oil. The
integrated cleanout system was successful in a
well with low BHP, large internal-diameter
completion and a long, undulating, horizontal
wellbore. ConocoPhillips and Schlumberger plan
to continue using the system to help improve
cleanout efciency on other difcult wells in the
Kuparuk River Unit.
Improving Post-Stimulation Cleanout Efciency
As operators develop more low-permeability
reservoirs, hydraulic-fracturing stimulation of
highly deviated or otherwise complex horizontal
wellbores has become a relatively standard
practice. After fracturing, however, varying
amounts of proppant are left behind and must be
removed before production begins (see New
Fibers for Hydraulic Fracturing, page 34).
Since 1996, hydraulic fracturing in the Valhall
eld, offshore Norway, has become the preferred
Wait on rig
and other delays
Weather 8%
delays
16%
Nonproductive
time
6%
NORWAY
Stavanger
Valhall field
Hod field DENMARK
GERMANY
method of stimulation by operator BP, formerly
Amoco. In the North Sea, the cost of CT
operations is high and often requires a
stimulation vessel and a team of more than 20
completions and operations specialists. With
proppant-cleanout operations accounting for
about 35% of CT utilization time in the Valhall
field, improving the efficiency of cleanout
operations would not only reduce cost, but would
also bring wells on line faster, generating
incremental production revenue (below).11
9. Slick-water refers to a water-base uid with additives
designed to reduce friction pressure. Slick-diesel refers
to an oil-base uid with friction-reducing additives.
10. A gas hydrate is a solid crystalline structure consisting of
water with gas molecules in an ice-like cage congura-
tion. Water molecules form a lattice structure into which
many types of gas molecules can t. Most gases, except
hydrogen and helium, can form hydrates.
11. Zhou W, Amaravadi S and Roedsjoe M: Valhall Field
Coiled Tubing Post-Frac Proppant Cleanout Process
Optimization, paper SPE 94131, presented at the
SPE/ICoTA Coiled Tubing Conference and Exhibition,
The Woodlands, Texas, April 1213, 2005.
Final Start
rigdown rigup
Liner
3% 8% operations
9%
Hydraulic
fracturing
15%
Proppant cleanout
35%
11
 During 2004, engineers at BP and
Schlumberger built a database and documented
the CT cleanout processes used during 29 runs in
four completions. Each step in the cleanout
process was benchmarked with 24 parameters
including proppant properties, start depth,
penetration speed and rates, sweep-range
depths, circulation rate, time at bottom, pulling
out of hole (POOH) rate and time consumed on
each step. Of these parameters, engineers
focused on optimizing total effective time (TET),
dened as the sum of penetrating time, time
circulating bottoms-up and time washing from
bottom to surface.
Using PowerCLEAN software modules,
engineers analyzed previous cleanout operations
and defined opportunities for improving
efciency. Of particular note was the nding that
residual fracturing proppant appeared in the
wellbore in varying distribution patterns,
requiring that each of the design elements had to
be optimized for each specic wellbore section.
As a part of the optimization process,
engineers verified that a simple seawater
cleanout uid, already in use, provided sufcient
carrying capacity for single-run cleanouts.
Further analysis, modeling and simulations using
the PowerCLEAN software modules led to
recommendations for maximum CT speed when
penetrating ll or beginning the process of ll
removal. Specic parameters, such as whether
solids formed a bed on the low side of a wellbore
and the most efcient bite length into or out of
the ll, helped determine nozzle selection, ow
rates and uid rheology requirements.
The new design and recommendations helped
engineers optimize circulation rate and select
proper nozzles for each application. They also
were able to determine cleanout-uid rheology
requirements, calculate running speeds and bite
increments, and minimize or eliminate time on
bottom circulating bottoms-up. Speeds as fast as
20 m/min [66 ft/min] in the liner and tubing
sections were obtained while sweeping out
of the hole.
For BP, the Valhall proppant-cleanout
optimization project achieved project goals by
improving operational efciency and reliability,
and by reducing stuck-pipe risk. A total of 22 runs
across three completions used the PowerCLEAN
integrated cleanout system. The average TET
was reduced from 17.6 h/run to a new average
of 11.1 h/run (below). A savings of 6.5 h/run
represents a 37.2% reduction in average effective
cleanout time and indicates a significant
improvement in performance efciency.
Improving Cleanout Efciency in
Mature Fields
Located about 170 km [105.6 miles] northeast of
Kemaman, Terengganu, offshore Malaysia in the
South China Sea, the Dulang field began
production in the early 1980s. Operated by
PETRONAS, the eld comprises four platforms,
each with 15 to 22 wells. As in many maturing oil
elds, maintaining production rate in the Dulang
eld is a daunting task.
Although oil and gas wells in the Dulang eld
experience wax deposition, scaling and high
water cut, sand production remains the primary
cause of production decline. In 2004, at least
eight wells were shut in because of sand ll,
while production slowly declined in many others.
Wells in the Dulang field often require
intervention due to sand production every three
to six months. For PETRONAS, the speed and
efficiency of borehole cleanout operations
directly affect field production, revenue and
return on investment.
Large casing sizes, highly deviated wellbores,
elevated borehole temperature, low reservoir
pressure and limited production-platform deck
space all challenged the efciency of wellbore
cleanout operations. Early in 2004, PETRONAS
and Schlumberger engineers evaluated eight

Valhall Proppant Cleanout Times

35

30
Effective cleanout time, h

25

20
Average = 17.6 h/run
Average = 11.1 h/run
15

10

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55
Runs

1 2 3 4 A B C
Completions

> Improvement in cleanout efciency. Evaluation of CT cleanout runs in four completions (1, 2, 3 and 4)
in the Valhall eld (blue) indicated the average run time to be 17.6 hours. After applying the PowerCLEAN
integrated system, engineers reduced the average time over three completions (A, B and C) to 11.1 hours
per run (green), saving BP signicant time and cost, while getting wells back into production faster.

12 Oileld Review
 C-18L Trajectory

Number Well Treatment Depth at Deviation, Number Tubing, in. Casing, in. 1,000

True vertical depth, ft

top of fill, ft deg of runs
1 B-22L Sand cleanout 2,986 71 1 2 78 9 58
2,000
2 B-11L Sand cleanout 6,108 60 1 2 78 9 58
3 B-16 Wax cleanout N/A 80 1 2 78 3,000
4 C-22L Sand cleanout 3,035 75 2 2 78 9 58
5 C-9L Sand cleanout 4,954 50 1 2 78 9 58 4,000
6 C-17S Sand cleanout 7,888 70 1 3 12
-2,00 0
7 C-18L Sand cleanout 6,677 63 1 2 78 9 58 0 0
0
Nort 2,00 ft
8 D09L Sand cleanout 6,309 50 1 2 78 9 58 h, ft 2,000 4,000 East
,

> Improving cleanout efciency in a mature eld. In the South China Sea, PETRONAS has been operating the Dulang eld for more than 25 years. The
average wellbore deviation is 65 degrees, making cleanout operations difcult. Eight wells with trajectories similar to Well C-18L (right) were evaluated
as candidates for efciency improvement using the PowerCLEAN integrated system (left).

wells for sand and wax cleanout utilizing the
PowerCLEAN integrated systems approach
(above). Using the CoilCADE wellbore cleanout
module, engineers developed unique treatment
solutions for each of the eight wells. Cleanout
uids varied from gel and water to a combination
of nitried seawater and wax solvent, and were
designed for specic borehole conditions and
well congurations.
To restore and potentially enhance oil
production, engineers needed to clear the
wellbores of sand and debris, thus allowing
conveyance of slickline reservoir evaluation
tools. Then, each well could be evaluated,
stimulated if necessary and brought back on line
in a minimal amount of time.
Most wells in the field are similar, with
borehole deviations of approximately 63 degrees
and bottomhole temperatures (BHT) of 180 to
250F [82 to 121C]. Depending on design
requirements, engineers optimized uid cost on
several wells by selecting two different cleanout
fluid systems, an HEC-base fluid for tubing
cleanout and the PowerCLEAN uid system to
remove sand from the larger, and more difcult
to clean, tubing-to-casing annular space.
With the exception of Well C-22L, all cleanout
jobs were performed in one pass. Each treatment
was evaluated by slickline to confirm the
effectiveness of sand removal. On several wells,
engineers modied the design by switching to
nitrified foam fluids to compensate for lost
circulation and leaking completion tubing.
The integrated job design improved efciency uids, while mechanical engineers and uid-
and reduced time in hole by optimizing pump mechanics specialists develop nozzle technology;
rates, defining sand-bite sequences, properly the PowerCLEAN integrated wellbore cleanout
selecting nozzles for sand mobilization and system exemplifies this type of multidisci-
suspension, and accurately estimating chemical plinary collaboration.
consumption. Production was restored in seven Engineers have the tools and computing
of the eight wells immediately following support to quickly model, perform multiple itera-
treatment, while the other came back on line tions and optimize cleanout system performance
following acid stimulation. for most wellbore conditions and requirements.
On average, the PowerCLEAN integrated The successful integration of cleanout processes
systems approach to borehole cleanout reduced is helping many operators keep oil flowing
time in hole by 75%. The average job time was from their elds. This basic understanding of
reduced from two days to around one half-day per interdependent processes will lead the way
treatment. The operator saved time, improved to many more efficiency improvements in
return on investment, and returned the wells exploration and production systems. DW
to production at a much faster rate, realizing
as much as 900 bbl [143 m3] of incremental oil
per day.
Process Efciency
Efciency is essential in optimizing production
from aging oil elds and reservoirs that are
difficult to produce. By understanding the
interrelationships and potential synergies in
process elements, new technologies emerge,
helping operators return wells to production
faster. As nonproductive time decreases, costs
decrease and eld output increases.
Understanding key process elements is not
always straightforward, and often requires the
insights of experts from diverse disciplines. For
example, chemists generally develop cleanout

Summer 2005 13
 Spectroscopy: The Key to Rapid, Reliable
Petrophysical Answers
Well-completion decisions require fast, objective and reliable log interpretation.
A new, nearly automatic method for processing data from modern spectroscopy and
conventional logging tools gives operators that information quickly. An extensive
analysis of the relationship between rock properties and elemental concentrations
in core samples provided a reliable basis for this new service.
Dan Barson Petrophysical interpretations comprising at
Rod Christensen least porosity and water saturation are essential
OILEXCO Incorporated for decisions about acquiring pressure data, col-
Calgary, Alberta, Canada lecting fluid samples, running casing and
completing wells. Thus, providing reliable
Eric Decoster
answers in real time or within a few hours after
Caracas, Venezuela
logging is of utmost importance for operators.
Jim Grau Although reservoir characterization studies
Michael Herron involve more data and time to fine-tune the
Susan Herron interpretation for a particular reservoir, time
Ridgeeld, Connecticut, USA and data are always in short supply. Even for
these larger studies, a fast, reliable evaluation is
Udit Kumar Guru useful as a starting point and as a convenient
Cairo, Egypt summary of the logs.
Past attempts to provide a generalized inter-
Martn Jordn pretation package have been plagued by the
Petrleos de Venezuela S.A. (PDVSA)
need to manually define numerous parameters
Barinas, Venezuela
and formation-zonation levels. Ideally, these are
Thomas M. Maher selected by a skilled interpreter, or by referenc-
Apache Egypt Companies ing an established local database for the
Cairo, Egypt reservoir or formation. Unfortunately, neither of
these options may be available at the time
Erik Rylander required. An alternative approach is to deter-
Clamart, France mine many of these parameters automatically.
Now, elemental-concentration logs and the
Jim White SpectroLith lithology processing of spectra from
Aberdeen, Scotland neutron-induced gamma ray spectroscopy tools
make it possible to estimate all matrix parame-
For help in preparation of this article, thanks to Bill Batzer
and Lisa Stewart, Ridgeeld, Connecticut, USA; Matt Garber, ters automatically with at least the same
Cambridge, England; Martin Isaacs, Sugar Land, Texas, accuracy as conventional, time-consuming tech-
USA; Daniel Valois, Barinas, Venezuela; and
Richard Woodhouse, consultant, Surrey, England. niques. The number of parameters is
DecisionXpress, ECS (Elemental Capture Spectroscopy), dramatically reduced, in the optimal case to just
ELANPlus, GLOBAL, GLT (Geochemical Logging Tool),
Litho-Density, MDT (Modular Formation Dynamics Tester), one: the formation water resistivity, Rw. At the
Minitron, Platform Express, RST (Reservoir Saturation Tool)
and SpectroLith are marks of Schlumberger.
14
same time, more rigorously scientic conductiv-
ity and permeability models improve the
reliability of the results.
The combination of these techniques, known
as the DecisionXpress petrophysical evaluation
system, has been applied successfully in a wide
variety of siliciclastic reservoirs. For the time
being, this system is not applicable to carbonate
reservoirs, mainly because of the lack of good
scientific models and the difficulty of distin-
guishing calcite from dolomite in the presence
of gas. This article explains the basis for the
algorithms that allow the DecisionXpress system
to be fast, accurate and reliable, and shows
examples from several different environments,
such as Egypt, Venezuela and the North Sea.
First, however, it is useful to consider the limita-
tions of conventional techniques.
The Log-Evaluation Problem
A large part of well-log evaluation involves volu-
metric analysis. If the porosity and fluid
saturations are known, the determination of
uid volumes is trivial. Matrix permeability can-
not be measured directly by static log
measurements, but it can be estimated from
fluid and mineral volumes. The difficulty with
volumetric analysis is that there are far more
unknowns than measurements. In addition to
gas, oil and waterwhich can vary widely in
composition, density and relative abundance
from top to bottom of a hydrocarbon column
Oileld Review
there are many possible mineral components.
The log analyst also wants to know uid mobility,
for example whether the water in the formation
is irreducible or is free to produce.1
Modern logging suites can provide hundreds
of measurements, but these measurements are
not all independent. For example, many mea-
surements respond strongly to porosity, but none
uniquely identies the volume of oil. Confronted
with this challenge, the log interpreter is
obliged to work with models that reduce the
number of unknowns to an appropriate quantity
for the measurements available. For example, if
a reservoir is known to be a sandstone contain-
ing oil, the interpreter can exclude anhydrite
and gas from the model. The interpreter must
choose the model, so human intervention is
required from the very beginning. Although this
Summer 2005
choice may be simple for an adequately devel-
oped reservoir, it is likely to be difficult in an
exploration or appraisal well, or when the set of
measurements changes from those obtained in
offset wells that were used to establish the model.
Models contain parameters expressing the
response of the measurements to their compo-
nents. Some parameters are precisely defined,
for example the density of calcite. Some
responses vary widely, such as the gamma ray
response to shale. At this point, log interpreta-
tion programs take different approaches. Those
that emphasize ease and speed of use employ
simple models and allow only a few of the most
variable parameters to be set by an interpreter.
Those emphasizing accuracy offer complex mod-
els and allow most parameters to be modied by
the user (see A History of Log Interpretation
Methods, next page).
Whatever the approach, the most difficult
parameters to select are invariably those of clay
minerals. Clay type, volume and distribution
strongly affect the determination of porosity
from porosity logs, such as neutron, density and
sonic logs, and of water saturation from resistiv-
ity logs. In conventional log interpretation based
on a triple-combo logging suiteresistivity,
density porosity, neutron porosity, gamma ray
and spontaneous potentialthe volume of clay
is determined mainly from the gamma ray
response and the neutron and density measure-
ments. The subjectivity of gamma ray
(continued on page 18)
1. Irreducible water saturation is the lowest water satura-
tion, Swirr, that can be achieved in a core plug by
displacing the water with oil or gas. This state is usually
achieved by owing oil or gas through a water-saturated
sample, or by spinning it in a centrifuge to displace the
water with oil or gas.
15
 A History of Log Interpretation Methods

Log interpretation techniques have progressed

1940s to 1950s
from the linear solutions of simple equations
Charts and Nomograms
in the 1940s to todays mathematical inver-
Step-by-step, manual process Few parameters
sions and neural networks (right).1 The Simple models
development has been driven not only by
improvements in computer technology, but
also by the increasing number of well-log mea-
surements and the improved understanding of
Overlays
log responses. The main goals of log inter- Curves presented on selected Simple models
pretersto determine porosity, water scales and read with Few or no parameters
saturation and permeabilityhave not transparent rulers
1960s to 1970s

changed. What has changed is our ability to

estimate these quantities more quickly and
reliably in a wider range of formations, and to
Sequential or Deterministic Methods
compute other outputs, such as irreducible
Complex logic with iterative Explicit parameters:
water saturation and mineralogy. loops few or many
The foundation of quantitative log interpre- Limited model flexibility
tation is the set of relationships introduced by
G. E. Archie in 1941.2 At that early stage,
interpretation was a sequential processrst
Simultaneous or Statistical Methods
1980s to 1990s

determine porosity from a sonic, neutron or

density log, and then nd water saturation Constrained inversion through Usually many explicit
minimization of uncertainty parameters
using the resistivity log. This process was Greater flexibility in model
accomplished with charts and nomograms,
which became increasingly complicated as
more porosity logs became available and as
the effects of clay and of the invaded-zone Neural Network
fluids were recognized and quantified. Log Inputs mapped to outputs Minimal number
based on training datasets of parameters
interpretation was no longer a simple Implicit model (inputs, outputs)
sequential process, but one with many
1990s to 2000s

options and iterations.

Such iterations were no problem for the cal-
culators and computers being introduced at DecisionXpress System
that time. By the end of the 1960s, complex Nearly automatic lithology and Implicit model
programs, such as the SARABAND system, hence other outputs Minimal number of
parameters
could use all existing log measurements, esti-
mate clay volume from a variety of sources
and calculate the uid saturations in both the > The development of interpretation methods for multiple logging tools since the 1940s.
invaded and uninvaded zones.3 These pro- Single logging-tool interpretation is not shown.
grams operated sequentially, for example rst
estimating clay volume, then porosity and
nally water saturation, but iterated exten-
sively to rene the answer. Programs were
tailored for particular types of formations, for

16 Oileld Review
 example shaly sands, and for particular log-
ging measurements and response equations.
However, the complexity of the logic made it
increasingly difcult to add a new measure-
ment or interpretation idea later.
In the late 1970s, the idea of treating log
interpretation as a problem of mathematical
inversion was introduced.4 Each measurement
has a response equation that can be expressed
as a set of unknown formation volumes, each
multiplied by a parameter. When there are at
least as many equations as unknown volumes,
the latter can be found by common inversion
methods. The solution can be constrained, for
example by not allowing the porosity to
exceed a specied amount, and each response
equation can be given a different weight. In
this way, the logic in the sequential programs
could be mimicked, but it was not necessary
to rewrite the software to add or subtract a
measurement or a model.
In the 1980s, inversion methods were fur-
ther developed and their computing time
reduced to allow different models to be run
simultaneously.5 The most appropriate model
could then be selected for each interval, either
manually or using some automatic criterion.
Whatever method is selected, the main
tasks of computerized interpretation have
remained the same. First, the input logs need
to be edited, depth matched and environmen-
tally corrected. These tasks are increasingly
handled during acquisition, but still remain a
concern in difcult conditions such as
1. For a detailed review: Marett G and Kimminau S: Logs,
Charts, and Computers: The History of Log Interpreta-
tion Modeling, The Log Analyst 31, no. 6
(NovemberDecember 1990): 335354.
2. Archie GE: The Electrical Resistivity Log as an Aid in
Determining Some Reservoir Characteristics,
Transactions of the American Institute of Mining and
Metallurgical Engineers, 146. New York: American
Institute of Mining and Metallurgical Engineers (1941):
5462.
3. Introduced by Schlumberger in 1970, the SARABAND
system was the rst computerized reservoir analysis.
For more information: Poupon et al, reference 2, main text.
Summer 2005
washed-out boreholes. Second, the parame-
ters and, where the option is available, the
formation model, have to be chosen. Last,
results must be checked for quality and the
parameters or model changed until the inter-
pretation is satisfactory.
Parameter selection has always been a key
subject in log interpretation. Manual parameter
selection relies on selecting values from mea-
surements (like mud-ltrate properties), logs,
crossplots or histograms, for example by looking
for the apparent water resistivity in a water
zone. Unfortunately, there is no certainty that
an interval is water-bearingthat is an inter-
pretation in itself. Manual parameter selection
is often therefore a matter of judgment.
Most automatic parameter-selection meth-
ods implement the logic behind a manual
method, with the same limitations. In certain
cases, inversion can determine parameters by
making use of the fact that the parameters are
constant over an interval. Finally, parameters
can be chosen from databases specic to a
particular reservoir, formation, geographical
area or geological environment. These
databases range from simple Rw tables to sets
of procedures and the experiences of experts.
Quality control is even more subjective than
parameter selection. Reconstructed logs
those computed from the solution and the
parameters and model usedshow whether
the solution respects the input logs, but do
not indicate whether the parameters or model
are correct. In practice, the quality of the
4. Mayer C and Sibbit A: GLOBAL, a New Approach to
Computer-Processed Log Interpretation, paper SPE
9341, presented at the SPE Annual Technical Confer-
ence and Exhibition, Dallas, September 2124, 1980.
5. Quirein J, Kimminau S, LaVigne J, Singer J and Wendel F:
A Coherent Framework for Developing and Applying
Multiple Formation Evaluation Models, Transactions
of the SPWLA 27th Annual Logging Symposium, Hous-
ton, June 913, 1986, paper DD.
result depends on the interpreters judgment
and comparison with other data, such as core
analysis, well tests and production results.
Experienced interpreters do not use software
to nd the solution, but rather to implement
and rene ideas the interpreter gleans from
studying the raw logs. However, this experi-
ence does not have to be general and take a
long time to developit can be obtained
quickly in specic reservoirs or areas.
More recent techniques minimize the prob-
lem of parameter selection. Articial neural
networks are trained to convert logs into
results on wells where results are already
knowneffectively nding internally the nec-
essary transforms and parameters for the
specic model and wells concerned. Once
trained, the networks can be applied nearly
automatically to other wells in which the
same model applies. Although neural networks
are most commonly used for lithology classica-
tion and for cases in which explicit transforms
are not well-known, for example permeability
estimation and reduced logging sets, they are
also applied to volumetric analysis.
Finally, the DecisionXpress system uses new
measurements that allow some or all of the
petrophysical properties to be determined
nearly automatically. It is unlikely to displace
other methods for detailed studies requiring
high levels of accuracy and exibility. It
should, however, provide a signicant improve-
ment for initial, rapid decision making. JS
17
interpretation is widely known and can be illus- not remove the need for accurate parameter An Express Solution
trated on almost any log example (below). selections. In other cases, the choice of models During the last 20 years, new logging measure-
Various techniques have been used to and parameters is often facilitated by using a ments have advanced interpretation. These
improve the estimation of clay volume. Some database of knowledge about a particular reser- improvements can be divided into two types
interpretation software uses the minimum of the voir, local area or type of geological environment those focused on a better definition of the
clay volumes estimated by different methods, to reduce the choices considerably and to mini- fluids and those focused on a better definition
based on the reasoning that errors in each mize the need for human intervention. However, of the solids.
method always cause an overestimate. 2 This such databases might not be available until after Direct fluid definition has been vastly
approach can minimize gross errors, but does an area has been developed. enhanced by developments in nuclear magnetic
resonance (NMR) logging tools. Since the main
properties of interestporosity, saturation and
permeabilityare fluid-related, NMR may
Invaded Formation appear to be the best option. There are limita-
Resistivity
tions, however, particularly with saturation
0.2 ohm.m 2,000
Apparent Resistivity 1
interpretations, because measurements are
0.2 ohm.m 2,000 made in the invaded zone, close to the borehole,
Mudcake
Apparent Resistivity 2 and because the NMR oil and water signals are
Washout 0.2 ohm.m 2,000 sometimes not clearly separated.
Bit Size Apparent Resistivity 3 The other option is to dene the volumes of
6 in. 16 0.2 ohm.m 2,000 solids and then apply familiar equations to
Caliper Apparent Resistivity 4 determine the main reservoir properties from
6 in. 16 0.2 ohm.m 2,000
other measurements. For example, porosity can
Hole Deviation Apparent Resistivity 5 Photoelectric Factor
be determined accurately from the density log if
-10 deg 90 0.2 ohm.m 2,000 0 10
the matrix density is known. Water saturation
Cable Spontaneous Potential Invaded Zone Resistivity Thermal Neutron Porosity
Tension -430 mV -350 0.2 ohm.m 2,000 45
can be estimated from resistivity if clay conduc-
% -15
0 lbf 6,000 Gamma Ray True Formation Resistivity Formation Density tivity and distribution are known.
MD, ft 0 gAPI 200 0.2 ohm.m 2,000 1.95 g/cm3 2.95 The DecisionXpress system follows this sec-
ond option.3 Its solution is based on measuring
Shale the concentration of some of the elements in
rocks and then estimating the major matrix
6,250 properties from these concentrations. Measuring
elemental concentrations is not new: chemical
elements have been detected with pulsed neu-
tron spectroscopy logging tools since the late
1970s, and concentrations were specifically
derived for openhole formation evaluation by the
GLT Geochemical Logging Tool in the mid-
??? 1980s.4 Unfortunately, the GLT system was not
widely used for several reasons: The GLT tool-
6,300
string was long, operations were slow and
therefore costly, the tool was not combinable,
and the interpretation was complex. The
recently introduced ECS Elemental Capture
Spectroscopy sonde is short, simple to use and
fast to run, and it measures sufcient elements
to evaluate the lithology (next page, top).
At each depth level, the processing ows lin-
early, starting with the computation of lithology,
???
including clay volume, and proceeding through
> Log of a siliciclastic sequence, illustrating some of the grain density, porosity, permeability and
difculties of gamma ray interpretation. The gamma ray log, saturations (next page, bottom). The entire com-
neutron-density separation and resistivity all clearly indicate putation can be performed in real time, while
shale above 6,246 ft [1,904 m] measured depth (MD). However logging, and in most situations with the selec-
at 6,296 ft [1,909 m] and below 6,348 ft [1,935 m], the gamma ray tion of only one parameter, Rw, which is often
log indicates shale, but the other logs do not. Also, the minimum
gamma ray reading of 30 gAPI may or may not indicate pure known in developed reservoirs. Because the out-
clay-free quartz. put provides all log-derived parameters needed

18 Oileld Review
 1986 1991 1996 < Spectroscopy logging tools for lithology deter-
GLT Geochemical RST Reservoir ECS Elemental Capture mination. Each tool has at least one source that
Logging Tool Saturation Tool Spectroscopy emits high-energy neutrons into the formation,
Sonde and one detector that measures the gamma
rays emitted by the reactions of neutrons with
elements in the formation. The early GLT
Geochemical Logging Tool also incorporated nat-
ural spectral gamma ray and aluminum activation
measurements; it was long and slow to log, and it
Telemetry cartridge was not combinable with conventional logging
tools. The RST Reservoir Saturation Tool was
designed for cased-hole evaluation, and can
also provide inputs to the SpectroLith technique.
Detector
acquisition The ECS Elemental Capture Spectroscopy sonde
cartridge is the optimal spectroscopy tool for openhole
lithology and matrix property determination
using the SpectroLith technique and other
Far associated techniques.
Neutron detectors detector
Low-energy
neutron source

Neutron detectors
Near AmBe
detector source

Detector 2. Poupon A, Clavier C, Dumanoir J, Gaymard R and Misk A:

Minitron Log Analysis of Sand-Shale SequencesA Systematic
source Boron sleeve Approach, Journal of Petroleum Technology 22, no. 7
(July 1970): 867881.
3. Herron MM, Herron SL, Grau JA, Seleznev NV, Phillips J,
Electronics
El Sherif A, Farag S, Horkowitz JP, Neville TJ and Hsu K:
Dewar flask Real-Time Petrophysical Analysis in Siliciclastics from
Heat sink the Integration of Spectroscopy and Triple-Combo Log-
ging, paper SPE 77631, presented at the SPE Annual
Technical Conference and Exhibition, San Antonio, Texas,
USA, September 29October 2, 2002.
70 ft long 36 ft long 15 ft long 4. Hertzog R, Colson L, Seeman O, OBrien M, Scott H,
6 sondes Through tubing 5-in. OD (with McKeon D, Wraight P, Grau J, Ellis D, Schweitzer J and
Herron M: Geochemical Logging with Spectroscopy Tools,
2 sources [111/16-in. or boron sleeve)
SPE Formation Evaluation 4, no. 2 (June 1989): 153162.
2 passes 21/2-in. outside Chemical
<600 ft/h diameter (OD)] source
Minitron source 1,800 ft/h
<200 ft/h

Total porosity, T

Permeability, k Irreducible
water saturation
Relative
SpectroLith Effect of clay on permeabilities
lithology conductivity, Q v and water cut
Elemental
concentration logs Water saturation, Sw
(Si, Ca, Fe, S, Gd, Ti)
Density and Reservoir
neutron matrix Total porosity, T volumes
properties
Density, neutron and
resistivity logs

User-selected Switches for anhydrite Clay cation Formation water Cutoffs for
parameters: and feldspar level exchange capacity salinity permeabilities
and water cut
> Processing ow and user-selectable parameters in DecisionXpress processing. Blue rectangles represent input data,
green rectangles represent output data, and yellow rectangles indicate intermediate computations. The
straightforward nature of the process contributes to its robustness.

Summer 2005 19
 for picking points to measure pressure, fluid-
Porosity
sampling intervals, and sidewall coring
Siderite
locations, it is crucial in making completion
Clay-Bound Water Anhydrite decisions (left). Some conventional analysis
Capillary-Bound Water Pyrite packages can also claim speed and automation
Water
Free Water Carbonate once the analysis is adapted for specic environ-
Pay Hydrocarbon ments. The key difference of this new technique
Water Hydrocarbon Quartz/ Feldspar/ Mica
Cable DecisionXpress Water Cut is that it gives an accurate and reliable result in
Tension Mineralogy Hydrocarbon 1 0 Moved Hydrocarbon Clay
most siliciclastic reservoirs anywhere in the
1 lbf 5,000
Gamma Ray Flow Profile Intrinsic Permeability Porosity Volume world. To justify this claim, we will examine the
L P K SR
MD, ft 0 gAPI 200 0 1 10,000 mD 0.1 50 % 0 0 % 100
basis on which its algorithms are built.
Elemental concentrationsSpectroscopy
tools like the ECS sonde actually measure a
gamma ray spectrum, or the number of gamma
6,250 rays received by the detector for each energy
level. The gamma rays are produced when high-
energy neutronsfrom a minitron or from a
radioactive source such as americium [Am] and
beryllium [Be]bombard the formation and
lose energy through scattering, primarily by
hydrogen. When slowed down to thermal energy,
a neutron that collides with the nucleus of cer-
tain atoms can be captured; in this process, the
nucleus is excited, and it emits gamma rays with
6,300
a distribution of energies that is characteristic
of the element. These gamma rays may be
degraded by scattering in the formation and the
detector, but there is sufcient character in the
nal spectrum to recognize the peaks caused by
different elements (next page, top).
The next step is to calculate the proportion,
or relative yield, of gamma rays due to each ele-
ment. To do this, the measured spectrum is
6,350 compared with the standard spectrum acquired
by Schlumberger for each individual element in
> DecisionXpress output for the previous log (page 18 ). A light gray mask indicates log intervals in the Environmental Effects Calibration Facility in
which the input data are of poor quality because of hole conditions or other problems. The porosity, Houston. The spectrum is inverted to obtain the
permeability and uid saturations are summed and averaged over the pay interval using cutoffs on yields of the principal contributing elements.
permeability and water cut selected by the user. These can also be presented in a table. Quality- These include some of the most diagnostic and
control ags in the far right track indicate the interpretation description for lithology (L), porosity (P),
abundant elements in sedimentary rocks, in par-
permeability (K), saturation (S) and relative permeability (R); green indicates a favorable interpretation,
yellow means a moderately favorable interpretation and red reects an unfavorable interpretation. ticular silicon [Si], calcium [Ca], iron [Fe] and
The shale interval above 6,240 ft [1,902 m] is largely affected by poor hole conditions. sulfur [S]. Titanium [Ti] and gadolinium [Gd]
can also contribute signicant signal and there-
fore must be solved for, even though they are not
abundant elements. The yields of these six ele-
ments, all of which arise solely from the rock
matrix, are computed and used quantitatively in
further processing. Other elements, such as
5. Sedimentary minerals contain single or multiple oxides. by variable i. Once F has been calculated at each level, hydrogen [H] and chlorine [Cl], are also mea-
Even clay minerals can be treated as complex mixtures of the percent by dry weight, or elemental concentrations, sured but used only qualitatively.
oxides. Concentrations are expressed in percent by are computed from Wi = F * Yi / Si.
weight, because it is the mass and not the volume of an Herron SL: Method and Apparatus for Determining Ele-
element that contributes to the measured yield. mental Concentrations for Gamma Ray Spectroscopy
6. The oxide closure model as applied to the ECS tool can Tools, US Patent No. 5,471,057 (November 28, 1995).
be expressed as: F { Xi *Yi / Si } = 1, where F is the 7. Herron MM, Matteson A and Gustavson G: Dual-Range
unknown normalization factor, Y is the measured relative FT-IR Mineralogy and the Analysis of Sedimentary Forma-
yield, X is the known oxide association factor, and S is the tions, paper 9729, presented at the Annual Society of
known relative detection sensitivity. The summation is Core Analysts Conference, Calgary, September 710, 1997.
over the six measured rock-matrix elements, designated

20 Oileld Review
 The yields are only relative measures
because the total signal depends on the environ-
ment and varies throughout the logged interval.
To obtain the absolute elemental concentra-
tions, we need additional informationin this

Number of gamma rays detected, counts per second

case from the principle of oxide closure. This
principle states that a dry rock consists only of a
Gd
set of oxides, the sum of whose concentrations
must be unity.5 If we can measure the relative H
yield of all the oxides, we can calculate the total
yield and the factor needed to convert it to
unity. This normalization factor will then con- Si Fe
vert each relative yield to a dry weight
elemental concentration.
In practice, this process is not so straightfor- Cl
ward. First, we measure elements, not oxides,
but nature is helpful since the most abundant Inelastic
elements exist in only one common oxide, for
example SiO 2 for silicon. Thus, for most ele-
ments, an exact association factor supports
conversion of the concentration of the element 0 50 100 150 200 250
to the concentration of the oxide. Second, Gamma ray energy, measurement bin number
although the ECS tool measures a majority of > Typical gamma ray spectrum from the ECS tool in a siliciclastic
the most common elements, there are excep- environment that has no calcium or sulfur. The thermal neutron capture
tions, the most important being those of gamma rays are shown divided into the contributions of the different
potassium [K] and aluminum [Al]. Fortunately, elements present. Gamma rays from inelastic neutron reactions are also
present, but are not used quantitatively. The capture yields of iron [Fe]
the concentration of these elements is strongly and calcium [Ca] include small signals from aluminum and sodium. This
correlated to that of iron, so that they can be contamination is taken into account during further processing.
included in the oxide association factor for iron.6
The results have been validated by comparison
to chemical concentrations measured on core
samples (below right).
Elements to mineralsThe next step is to 200
convert elemental concentrations into mineral
groups. Earlier geochemical techniques were 300
designed to determine as many minerals as pos-
sible. In this DecisionXpress technique, the
400
primary goal is an accurate and reliable total
clay content or weight fraction, with remaining
500
minerals being divided into carbonates or into
Depth, ft

quartz, feldspars and micas (QFM). Develop-

600
ment of this technique was based on the study of
more than 400 core samples from different sand
and shaly sand environments. Each sample was 700

crushed, mixed and split into two fractions

one to determine elemental concentrations 800

through chemical analysis, the other to deter-

mine mineralogy using the Fourier transform 900
infrared (FT-IR) procedure.7 The mineral stan-
0 50 0 40 0 20 0 20 0 4 0 40
dards for the FT-IR procedure included 26 Silicon, % Calcium, % Iron + 0.14 Al, % Sulfur, % Titanium, % Gadolinium, ppm
by weight by weight by weight by weight by weight
minerals, all of which can be determined with
Elemental concentration, dry weight fraction
an accuracy that exceeds +/- 2% by weight.
> An example of the good agreement between six elemental concentrations measured on core (red)
and those derived by applying the oxide closure principle to the yields of the ECS tool (black). The iron
yield contains some signal from aluminum, so that it actually measures the iron concentration plus
14% of the aluminum concentration. The core data (red circles) are plotted using the same combination.

Summer 2005 21
 100 The study first examined the correlation
between total clay and a number of elements

Clay, % by weight
conceivably measurable with logs (left). Total
50 clay is the sum of the kaolinite, illite, smectite,
chlorite and glauconite fractions. In most wells,
aluminum gives the best correlation, which is
0 not surprising because clays are aluminosili-
0 10 20 0 5 10 0 2.5 5
Thorium, ppm Uranium, ppm Potassium, % by weight cates, and aluminum is an integral part of their
100 chemical composition. Potassium sometimes
correlates strongly when the dominant clay is
Clay, % by weight

illite, but the correlation is perturbed by potas-

50 sium in feldspars, micas and other minerals.
Thorium [Th], uranium [U], titanium [Ti] and
gadolinium [Gd] are trace elements that are
0
0 10 20 0 1 2 0 5 10 often enriched in shales, but these elements do
Aluminum, % by weight Titanium, % by weight Gadolinium, ppm not generally reveal a sufciently reliable corre-
100 lation for quantitative use, primarily due to
nonclay sources. Silicon shows a strong anticor-
Clay, % by weight

relation, decreasing from 46.8% by weight in

50
0
0 25 50 0 15
Silicon, % by weight Iron, % by weight
> Comparison of the concentrations of various elements measurable by logs
with measured clay concentration in one well. The top row contains
elements measured by natural gamma ray spectroscopy. The two trace
elements, Ti and Gd, and the three major elements measured by capture
gamma ray spectroscopy are shown in the middle and bottom rows, along
with aluminum, which is difcult to measure with wireline or logging-while-
drilling tools. As observed in many wells, there is a good correlation with
aluminum and a good anticorrelation with silicon. In this well, the
correlation with potassium is good, but this was an exception among the
wells studied, particularly at low clay concentration.
100
a b
Clay, % by weight
pure quartz to about 21% by weight in clays. Iron
is associated with heavy minerals, such as
siderite and pyrite, and the clay minerals illite,
30 0 20 40
chlorite and glauconite. Calcium occurs primar-
Calcium, % by weight ily in calcite and dolomite.
Aluminum is the best single elemental indi-
cator of clay, but it is difcult to measure in the
borehole. Because of its small capture cross sec-
tion, aluminum does not produce sufficient
capture gamma rays to make a statistically reli-
able measurement. In the past, aluminum was
measured by inducing neutron activation, a
technique that required complex hardware, such
as that in the GLT tool.
For this reason, researchers looked for other
methods with better statistical precision to esti-
mate total clay. The silicon anticorrelation is
good, but is disturbed by the presence of carbon-
c ate minerals, siderite and pyrite (left). These
minerals act like clay to reduce the amount of sil-

icon, but can be accounted for by measuring

50 calcium, iron and, when available, magnesium
[Mg], whose measurement is discussed below.
Thus, by combining four elementsSi, Ca, Fe
and Mgit is possible to nd a correlation with
0
0 50 100 0 50 100 0 50 100 total clay that has nearly the same slope in all
100SiO2 100SiO2 CaCO3 MgCO3 100SiO2 CaCO3 MgCO3 1.99 Fe wells, a small degree of scatter, and a near-zero
> Data from 12 wells illustrating how clay concentration is estimated from major intercept (next page, top). When examining these
elements. The measured clay concentration shows a clear trend with (100 SiO2) that is plots, it is important to focus on the clay-poor
disturbed mainly by carbonate minerals (a). When calcite and dolomite are subtracted region where reservoirs occurthe correlation in
from the previous estimate, the tight trend is disturbed only by siderite and pyrite (b). the shales is less important. With the exception of
When iron-rich minerals are also subtracted, the correlation is further improved,
showing how clay can be estimated from four elements (c). In practice, magnesium is Wells 11 and 12, discussed below, these results
not measured by capture gamma ray spectroscopy, but the interpretation provides a show a strong, unique correlation between ele-
total carbonate (calcite + dolomite) that effectively produces results identical to those mental concentrations and total clay in a wide
shown in the middle plot (b). For complete lithology interpretation, dolomite can be
range of siliciclastic reservoirs.
estimated from the photoelectric factor from the Litho-Density photoelectric density log
or Platform Express integrated wireline logging tool measurements.
8. Ellis DV: Well Logging for Earth Scientists. New York:
Elsevier (1987): 190.

22 Oileld Review
 At this point, it is worth examining the corre- 100
Well 1 Well 2 Well 3 Well 4
lation between measured total clay and the

Clay, % by weight
traditional total gamma ray on the same data
(below right).8 The gamma ray is calculated from 50

the sum of its contributing elementsK, Th and

Uand is therefore independent of porosity. As
expected, there is a general correlation. How- 0

ever, the slopes and offsets vary widely, and 100

Well 5 Well 6 Well 7 Well 8
there is often considerable scatter, particularly

Clay, % by weight
in comparison with the estimation based on ele-
mental concentrations. 50

Wells 1 and 2 illustrate the wide range in

slope. An extrapolation to pure clay would give a
gamma ray reading of 100 gAPI in Well 1, but 0

would give 500 gAPI in Well 2. Wells 4 and 12 100

Well 9 Well 10 Well 11 Well 12
illustrate the range in offsets, or zero clay read-
ings. An extrapolation to zero clay gives 30 gAPI Clay, % by weight
in Well 4 and 70 gAPI in Well 12. Such variations 50

are widely known and are partially circumvented

in practice by using local knowledge and cali-
0
brating the gamma ray to core data in a 0 50 100 0 50 100 0 50 100 0 50 100
particular reservoir. Estimated clay, % Estimated clay, % Estimated clay, % Estimated clay, %
Calibration would give good results in several > Comparison of measured clay concentration with the concentrations estimated by Si, Ca, Fe and Mg
of the wells. However, the results are still not in 12 wells. Except in Wells 4, 11 and 12, the slopes are nearly the same and pass through the origin
satisfactory in terms of scatter and dynamic with no offset. The overall correlation coefcient is 0.94, with a standard error of 6.9% by weight. In
reservoir rocks that contain less than 25% clay, the standard error is smaller. The clay content tends to
range. In Wells 3, 5, 7 and 9, the scatter at about be underestimated in the shales; this underestimation is corrected in the SpectroLith implementation.
20% by weight clay is such that even a calibrated
gamma ray would indicate clay percentages
varying from 0 to 40%. This amount of clay can 100
mean the difference between reservoir and non- Well 1 Well 2 Well 3 Well 4
Clay, % by weight

reservoir rock, and makes quantitative use

difcult. Wells 11 and 12 are examples of small 50
dynamic range.
Wells 11 and 12, and to a lesser extent Well 4,
contain feldspar-rich sands. Feldspars and micas 0
are aluminosilicates, like clays, and therefore 100
affect the silicon content. These sands are han- Well 5 Well 6 Well 7 Well 8
Clay, % by weight

dled by using a different slope and introducing

an offset in the clay estimator (below). The 50

100

0
Clay, % by weight

100
Well 9 Well 10 Well 11 Well 12
Clay, % by weight

50

50

0 0
0 50 100 0 100 200 0 100 200 0 100 200 0 100 200
Estimated clay, % Gamma ray, gAPI Gamma ray, gAPI Gamma ray, gAPI Gamma ray, gAPI
> Comparison of measured clay > Comparison of measured clay concentration with gamma ray in the same 12 wells as in the previous
concentration with the gure (above ). The gamma ray has been computed from the thorium [Th], uranium [U] and potassium
concentrations estimated by Si, Ca, [K] concentrations measured on the samples using the formula: gamma ray = 4Th + 8U + 16K, where
Fe and Mg in Well 4 (crosses) and Th and U are reported in parts per million (ppm), and K is in % by weight. This is equivalent to using a
Wells 11 and 12 (open circles) using gamma ray log normalized to the solid fraction, or porosity-free. The slopes and offsets vary widely
the equation for arkosic, or high from well to well. Even after allowing for these, the correlations are poorer than when estimating
feldspar, sands. The correlation is using Si, Ca, Fe and Mg, especially in reservoir rocks.
strong, particularly below 20% clay.

Summer 2005 23
current DecisionXpress implementation has
three different estimators corresponding to
arenite (feldspar content < 10%), subarkose
(feldspar content between 10 and 15%) and the
rare arkose (feldspar content > 25%). Arenite is
the default.
The carbonate fraction is determined from
the calcium concentration, initially assuming
that the carbonate is calcite. Dolomite can be
detected and quantified by comparing the
expected photoelectric factor (PEF) with the
measured PEF. 9 The fractions of halite, coal,
Inversion
(spectral stripping)
Oxide closure
siderite, anhydrite and pyrite are all mea-
sured using information in the different yields
(below). The remainder of the rock is considered
to consist of quartz, feldspar and mica (QFM).
The extensive core studies helped scientists
develop an accurate and reliable method of esti-
mating clay from elemental concentrations
without the need for user intervention. This pro-
cess is captured in the SpectroLith algorithm.10
One important advantage is that it uses concen-
trations of major elements, as opposed to trace
elements that can be easily affected by sediment
Induced gamma ray
spectra
Elemental relative yields
(Si, Ca, Fe, S, Gd, Ti, H, Cl
and other capture and
inelastic yields and tool
background)
Elemental concentrations
% by dry weight
(Si, Ca, Fe, S, Gd and Ti)
diagenesis, depositional environment or the spu-
rious introduction of small amounts of heavy
minerals. The results are demonstrably superior
to those from the gamma ray log even when the
gamma ray analysis is calibrated with core. Also,
unlike the lithology analysis using neutron and
density porosities, the results are independent
of uid type, volume and density.
Matrix properties and porosityIn conven-
tional log analysis, matrix density is either taken
as a constant based on local knowledge or is
derived from mineral modeling. The former is
likely to be approximated, leading to errors,
while the latter involves analyst input and con-
trol. An alternative approach is to estimate
matrix density directly from elements. As with
the lithology study, elemental concentrations
and matrix densities were available on a large
number of core samples, in this case more than
600. The goal was to find the best correlation
between the matrix density and a linear combi-
nation of elements. Although the algorithm is
empirical, its rationale is logical.11 Sandstone
matrix density is approximately equal to that of
silica [SiO2], but increases as the concentra-
tions of calcium-, iron- and sulfur-bearing
minerals increase. Iron-bearing minerals have a
particularly strong effect on density, as reected
in the high coefcient for iron. A separate algo-
rithm with different coefficients is used for
arkosic sandstones.
A similar analysis leads to an algorithm for
the matrix response of the neutron log. Knowing
the properties of the rock matrix and the uid
normally those of mud filtrateit is
straightforward to calculate the total porosity
from both neutron and density logs. In water
zones, the matrix-corrected porosities should
agree regardless of the volumes of clays or heavy

Mineral Element Used Comment

Anhydrite, CaSO4 Sulfur User elects to solve for either anhydrite or
pyrite. Corresponding weight % of Ca or Fe
SpectroLith model is subtracted from measured weight %
Pyrite, FeS2 Sulfur before calculating other lithologies.

Siderite, FeCO3 Iron From the iron remaining after computing

SpectroLith lithology
% by dry weight
(clay, carbonate, QFM
and special minerals)
pyrite and clay
Coal, CHaNbOc Hydrogen From excess hydrogen above average
hydrogen level in well. Other minerals
normalized to noncoal fraction

Properties such as T , b , Halite, NaCl Total count rate If detected, lithology set to
clay, carb and QFM above threshold 100% halite

> Overview of the SpectroLith algorithm. The processing ow (left ) starts

Lithology
Volume % of rock, with fluids
(clay, carbonate, QFM and
special minerals)
with capture yields and determination of lithology in % by dry weight. The
lithology is subsequently converted to % by volume using porosity, log bulk
density and the density of the mineral components. The table summarizes
the logic used to detect special minerals and coal (above).

24 Oileld Review
minerals. In gas zones, there should be a clear
crossover, unmasked by the effects of clay.
Finally, the total porosity, T, for use in further
computations, is taken as two-thirds of the den-
sity porosity, D, plus one-third of the neutron
porosity, N. This expression yields an approxi-
mate but reliable estimate of T for any
formation uid.
Water saturationMany equations are avail-
able to compute water saturation from resistivity.
Because we have a reliable measure of clay vol-
ume, it makes sense to select an equation that
uses clay volume explicitly and is based on labo-
ratory studies. The Waxman-Smits-Thomas
equation satisfies these conditions and is the
current choice in the DecisionXpress system.12
The Waxman-Smits-Thomas equation con-
tains the only two parameters that must be
selected by the userformation water conduc-
tivity, Cw, and the clay cation exchange capacity
(CEC). Formation salinities vary far too widely
for a fixed default to be satisfactory. The clay
CEC default is 0.1 meq/g, a good value for most
illites and chlorites, and a good value for most
clay mineral assemblages the researchers
encountered in sedimentary rocks; pure kaolin-
ite clays and pure smectite clays are not present
in the extensive database.
In a water-lled formation, with Sw = 1, the
same equation is used to calculate the water-
lled formation resistivity, Ro, and the apparent
formation-water resistivity, Rwa.
k- permeabilityPermeability is calcu-
lated by a method developed for siliciclastic
formations based on the lambda parameter, .13
The lambda parameter is a measure of the effec-
tive diameter of dynamically connected pores,
and, in the simplest pore geometries, it can be
approximated from the ratio of pore volume to
surface area. Further, at high permeability, the
permeability is proportional to 2/F, where F is
the Archie formation factor and equal to 1/ 2.
Combining these leads to an expression that is a
form of the Kozeny-Carman relation and similar
to many others in the literature:
k ~ m* / (S/ Vp ) 2 ,
where S is the pore surface area and Vp the pore
volume.14 The problem is then how to measure
the ratio S/Vp from logs, and how to adapt the
equation at low permeability. In the mineral
form of k- permeability, the ratio S/Vp is esti-
mated from the volumes of the minerals present.
This is possible by first removing the effect of
porosity in the ratio, leaving two terms, the
Summer 2005
10 4
Measured permeability, mD
10 2
10 0
k- estimate
10-2
0 10 20
Measured porosity, %
> Permeability calculation based on the lambda parameter, . Measured
porosity and permeability values (blue) and the k- estimate (red) for clay-
free Fontainebleau formation quartz arenites are shown at left. Measured
permeability (blue) versus k- estimate (red) for the same formation
appears at right. The correlation coefcient for the logarithms is 0.99.
matrix density and the specic surface area per
unit mass, S0.15 S0 is a characteristic of different
types of minerals. It is known that clays have
high S0 and make by far the largest contribution
to pore surface area in shaly sands. It has also
been observed that the total S0 in a rock can be
approximated by a linear combination of the
9. PEF refers to a log of photoelectric absorption properties.
The log measures the photoelectric absorption factor, Pe,
which is dened as (Z/10)3.6, where Z is the average
atomic number of the formation. Pe is unitless, but
because it is proportional to the photoelectric cross sec-
tion per electron, it is sometimes quoted in
barns/electron. Because uids have low atomic num-
bers, they have little inuence, so that Pe is a measure of
the rock matrix properties. The PEF of dolomite is less
than that of calcite. The PEF reconstructed from the com-
puted matrix fractions should equal the measured PEF if
the carbonate is pure calcite. If the measured PEF is less,
the difference is proportional to the fraction of dolomite.
See Hertzog et al, reference 4.
10. Herron SL and Herron MM: Quantitative Lithology: An
Application for Open and Cased Hole Spectroscopy,
Transactions of the SPWLA 37th Annual Logging Sympo-
sium, New Orleans, June 1619, 1996, paper E.
11. For arenites or subarkosic sandstones, researchers
found a single least-squares t with a correlation
coefcient of 0.97 and a standard error of 0.015 g/cm3
[0.936 lbm/ft3] as follows:
ma = 2.62 + 0.049 WSi + 0.2274 WCa + 1.993 WFe +
1.193 WS , where WSi , WCa , WFe and WS are the % by
dry weight of these elements.
Herron SL and Herron MM: Application of Nuclear
Spectroscopy Logs to the Derivation of Formation Matrix
Density, Transactions of the SPWLA 41st Annual Log-
ging Symposium, Dallas, June 47, 2000, paper JJ.
12. The Waxman-Smits equation for the conductivity
response of shaly formations is used to analyze core
data and to calculate water saturation from resistivity
and other logs. The model was developed by M. Waxman
and L. Smits with later contributions by E. C. Thomas. The
Waxman-Smits-Thomas equation may be stated as follows:
1/Rt = Ct = Tm* Swn* (Cw + BQv/Sw) ,
where Ct is conductivity, or the reciprocal of Rt, the mea-
sured log resistivity; Sw is the water saturation; m* is the
cementation exponent and is a well-dened function of
T and Qv ; n* is the saturation exponent set to 2; and Cw
10 4
Measured permeability, mD
10 2
10 0
10-2
k- estimate
10-4
30 10-4 100 104
Estimated permeability, mD
mass fractions of the minerals present.16 This
works well until the pore throats become blocked
at low porosity and permeability. Empirically, it is
found that when the initial k- estimate is less
than 100 mD, it must be reduced by a suitable
function. The quality of the k- estimates can be
judged from the examples (above).
is the formation water conductivity. The rst term is
equivalent to the Archie equation in clean formations.
The second term, BQv/Sw, represents the additional con-
ductivity due to clay, where B is a parameter that is a
function of temperature and Cw. Qv, the cation exchange
capacity (CEC) per unit pore volume, is directly related to
clay volume and its CEC. CEC is the quantity of positively
charged ions that a clay mineral or similar material can
accommodate on its negatively charged surface,
expressed as milli-ion equivalent per 100 g, or more com-
monly as milliequivalent (meq) per 100 g.
Smits LJM and Waxman MH: Electrical Conductivities
in Oil-Bearing Shaly Sands, Society of Petroleum
Engineers Journal 8, no. 2 (June 1968): 107122.
Waxman MH and Thomas EC: Electrical Conductivities
in Shaly Sands I. The Relation Between Hydrocarbon
Saturation and Resistivity Index; II. The Temperature
Coefcient of Electrical Conductivity, Journal of
Petroleum Technology 26, no. 2 (February 1974): 213225.
13. Herron MM, Johnson DL and Schwartz LM: A Robust
Permeability Estimator for Siliciclastics, paper SPE
49301, presented at the SPE Annual Technical Confer-
ence and Exhibition, New Orleans, September 2730, 1998.
14. Carman PC: Flow of Gases through Porous Media. Lon-
don: Butterworths Scientic Publications, 1956.
15. The pore surface area, S, within a bulk volume Vb, can be
written as a product of the specic surface area per unit
mass, S0, and the mass of the matrix, which is its volume,
(Vb - Vp), multiplied by its density, ma. The porosity, , is
given by Vb/Vp. Thus, S/Vp = S0 ma (1-)/.
16. On this basis, the initial k- estimate becomes:
k1 = 200,000 (m* +2) / {(1-)2 ma2 (60Wclay + 0.22Wsand +
2Wcarb + 0.1Wpyr)2} ,
where Wclay, Wsand, Wcarb and Wpyr are the weight frac-
tions derived previously, and the numerical coefcients
are derived by tting to experimental data. Theoretically,
and also in practice, this expression does not apply at
low permeability. When the initial k- estimate is less
than 100 mD, it must be reduced as follows:
k2 = 0.037325 k11.714 .
25

A F
R I C
A
East Bahariya
0 km 200
0 mi
> East Bahariya area, Egypt. Apache Egypt produces oil from
the Cretaceous sandstones of the Bahariya and Abu Roash
formations (right).
Irreducible water saturationTo judge
whether a reservoir will produce hydrocarbons,
water or a mixture of both, it is not sufcient to
know just the water saturation, Sw. A qualitative
judgment can be made by a simple comparison
of S w with the irreducible water saturation,
Swirr. If Sw is equal to Swirr, then there is no pro-
ducible water. More quantitatively, the effective
permeabilities of oil, water and gas can be esti-
mated using familiar relationships that depend
on Sw and Swirr. Irreducible water saturation is
therefore an important parameter. In
DecisionXpress processing, it is derived by the
Coates-Timur equation.17 This equation is nor-
mally used to estimate permeability, but may be
inverted to give Swirr using porosity and the k-
estimate of permeability:
Swirr = 100 2 / (100 2 + k0.5)
With information on lithology, porosity, water
saturation, permeability and irreducible water
saturation, the operator has most of the major
inputs needed to make reliable decisions. Let us
now see the results of applying this logic to vari-
ous sandstone reservoirs around the world.
26
Alexandria
Cairo
E G Y P T
200
Rapid Evaluation of Complex
Lithologies in Egypt
In the East Bahariya concession onshore Egypt,
Apache Egypt is drilling exploratory wells in Cre-
taceous sandstones of the Bahariya and Abu
Roash formations (above). Geological uncer-
tainty and abrupt changes in formation-water
resistivity make rapid petrophysical analysis at
the wellsite challenging but desirable. With two
drilling rigs in operation, timely decision making
is important to minimize the impact of these
uncertainties on operations.
The prospective Bahariya and Abu Roash
sandstones tend to be thinly bedded and vary
widely in grain size. Complex mineralogy, includ-
ing glauconite, complicates log interpretation.18
Apache expected that prompt, robust interpreta-
tions based on the DecisionXpress system would
. help geoscientists and engineers plan subse-
quent formation-evaluation operations, such as
formation testing and fluid sampling with the
MDT Modular Formation Dynamics Tester tool.
Apache selected the DecisionXpress service
in part because it integrates data from the Plat-
form Express and ECS tools to determine
mineralogy. This service also provides a continu-
ous matrix density measurement that can be
Western Desert and Nile Delta
Generalized Stratigraphy
Rock Unit
Age Average
Lithology Thickness, ft
Formation Unit
Khoman
A
1,250
A 400
B 350
C 200
Abu Roash
D 500
Cretaceous
E 600
F 200
G 1,000
Bahariya 950
Kharita 3,000
Alamein 160
Alam El Bueib 2,000
used in subsequent log processing. Apache vali-
dated the ECS-derived mineralogy of the
Bahariya and Abu Roash formations with side-
wall core analysis. The ECS data helped identify
zones containing signicant amounts of calcite.
This was not possible using the standard PEF logs,
which are affected by barite in the drilling mud.
In a recent East Bahariya exploration well,
EB-28, petrophysical evaluation using
DecisionXpress technology compared satisfacto-
rily with a conventional analysis carried out by
Apache (next page). On the basis of this inter-
pretation, Apache decided to run the MDT
device to better understand uid mobility and to
collect uid samples. The MDT permeability cor-
related well with permeability estimated
17. Timur A: Pulsed Nuclear Magnetic Resonance Studies
of Porosity, Movable Fluid, and Permeability of Sand-
stones, Journal of Petroleum Technology 21, no. 6 (June
1969): 775786.
Coates GR, Miller M, Gillen M and Henderson G: The
MRIL in Conoco 33-1: An Investigation of a New Mag-
netic Resonance Imaging Log, Transactions of the
SPWLA 32nd Annual Logging Symposium, Midland,
Texas, USA, June 1619, 1991, paper DD.
18. Glauconite is a silicate mineral found in sedimentary
rocks. It typically forms on continental shelves charac-
terized by slow sedimentation with organic matter
present in an oxidizing environment. In sufcient quan-
tity, it can form thick, sandy, green deposits.
Oileld Review
 Net Pay
Porosity
Net
Reservoir Clay-Bound Water Siderite
Mudcake Pyrite
Water Capillary-Bound Water
Perforated Intervals
Washout Hydrocarbon Free Water Carbonate
Bit Size
Water Water Cut Hydrocarbon Quartz/ Feldspar/Mica
6 in. 16 DecisionXpress 1 0
Caliper Mineralogy Hydrocarbon Moved Hydrocarbon Clay Hydrocarbon
Intrinsic
6 in. 16 Gamma Ray Flow Profile Permeability Porosity Volume Sw
L P KSR
MD, ft 0 gAPI 150 0 1 10,000 mD 0.1 50 % 0 0 % 100 100 % 0

X,600

X,650

X,700

> Real-time petrophysical analysis of an East Bahariya well. This standard presentation displays borehole and
depth information, red net pay ags and yellow net reservoir ags in the depth track. Track 1 shows lithology
from the Platform Express integrated wireline logging tool. Perforations in four zones and the ow prole are
shown in Track 2. However, the water cut information in Track 3 reveals a zone near X,675 ft that ultimately
produced water. Fluid interpretations in Track 4 suggest that the best oil potential exists just below X,600 ft and
around X,700 ft. The dominant minerals, shown in Track 5, include quartz, feldspar and mica (yellow) and clay
(gray) with minor amounts of carbonate minerals (blue). Summarized in quality-control Track 6 are lithology (L),
porosity (P), permeability (K), saturation (S) and relative permeability (R); green reects a favorable
interpretation, yellow means a moderately favorable interpretation and red indicates an unfavorable
interpretation. Track 8 shows hydrocarbon volumes.

Summer 2005 27
 using the DecisionXpress system (left). In
Net Pay
Porosity addition, production results confirmed the
Net Water
Reservoir Clay-Bound Water Siderite
DecisionXpress analysis.
Hydrocarbon Apache employed DecisionXpress and ECS
Mucake Capillary-Bound Water Pyrite
Water Cut technology in other recent exploration wells
Washout Free Water Carbonate
1 0 drilled in two other concessions in Egypt for
Bit Size Intrinsic
Water Hydrocarbon Quartz/ Feldspar/ Mica quick-look petrophysical evaluations to support
6 in. 16 DecisionXpress Permeability
Caliper Mineralogy Hydrocarbon 10,000 mD 0.1 Moved Hydrocarbon Clay Hydrocarbon wellsite decision making. Estimates of the ratio
6 in. 16 Gamma Ray Flow Profile Mobility Porosity Volume Sw of net pay to gross pay from the DecisionXpress
L PK SR computation matched net-to-gross computations
MD, ft 0 gAPI 150 0 1 10,000 mD 0.1 50 % 0 0 % 100 0 % 100
developed from time-consuming proprietary
petrophysical analysis. DecisionXpress answers
were typically available before production casing
was run, which helped the operator estimate the
X,400
value of exploratory wells and decide whether to
run casing.
Based on this success, plans are now being
made to use the DecisionXpress software in real
X,450 time to design more effective wireline pressure-
and fluid-sampling programs as well as to
provide faster quick-look petrophysical analyses.

Real-Time Decisions in Venezuela

X,500
The Guata eld is a mature oil eld operated
by Petrleos de Venezuela S.A. (PDVSA), located
in the state of Apure, near the Venezuela
Colombia border (next page, top). This eld pro-
X,550
duces light oil, between 28 and 32 dAPI,
primarily from the Guardulio and Arauca mem-
bers of the Guata formation.
In recent years, PDVSA initiated a sustained
high level of drilling activity in the Guata eld
X,600 to maintain high levels of production. With three
active drilling rigs operating more than 300 km
[186 mi] from PDVSA headquarters in Barinas,
the company sought a reliable method for inter-
preting logs quickly at the wellsite. The PDVSA
X,650
operations team in Barinas turned to induced
gamma ray spectroscopy with the ECS sonde and
the DecisionXpress system.
The Guafita formation is a sand-shale
X,700
sequence in which conventional log interpreta-
tion might be expected to be straightforward. In
reality, several factors complicate log interpreta-
tion. First, the Guafita formation is highly
resistivea clean sand will produce water at a
X,750 resistivity of 10 mS/m [100 ohm.m], and oil at
3.3 mS/m [300 ohm.m]. This is because the for-
mation connate water is unusually fresh, varying
> Permeability and mobility interpretations. The MDT device measured formation pressure and uid from a low of 100 parts per million (ppm) to a
mobility at nine depths within the Bahariya formation and at three depths in the overlying Abu Roash high of about 2,500 ppm of equivalent sodium
G zone (Track 3). Permeability calculated by real-time DecisionXpress processing (Track 3) closely chloride [NaCl]. Therefore, master calibration of
matches MDT uid mobilities.
the induction tool requires great care because
the difference between 3 and 10 mS/m is signi-
cant, given that induction tools respond to
conductivity, not resistivity.

28 Oileld Review
 This simplified resistivity approach is ade-
quate for the clean sands, but it is inadequate
when clay is present and surface conductivity
effects become signicant. At such low connate-
water salinities, conditions in Guafita are
SOUTH AMERIC A
outside the traditional range of application of
conventional saturation equations such as the
Waxman-Smits model.19 In addition, the use of
oil-base mud precludes acquiring a spontaneous
potential curve, which in turn prevents the log
analyst from using saturation equations
designed specifically for freshwater environ-
ments, such as the Sen-Goode-Sibbit equation.20
After analyzing this problem, PDVSA decided Caracas
to focus on reducing the uncertainty associated
with implementation of a conventional satura-
tion model in Guata. The company began with Guafita
a thorough analysis of produced waters from var- field
Apure VENEZUELA
ious intervals in several wells to optimize the Rw
value to be used in each geological interval.
At the same time, PDVSA recognized that a
traditional estimate of clay content, always
biased by the gamma ray log, tended to overesti-
mate the amount of clay present in the 0 km 300

formation and invalidated a saturation clay cor- 0 mi 300

rection. The high radioactivity often observed in

> Guata eld, Venezuela. Located in the Apure state near the
the Guata sands is usually attributed to incom-
VenezuelaColombia border, the eld was discovered in 1984
patibility between the original connate water
and produces from Miocene and Oligocene reservoirs of the
and the water from the active aquifer below, Guata formation.
which is fresh and likely to originate from mete-
oric recharge. As the aquifer rises, radioactive
350
salts deposited in the formation increase the
overall radioactivity, and lead to overestimation ay
V cl
of clay content. *
300
324
In discussions with Schlumberger, PDVSA iden- 8+
tied induced gamma ray spectroscopy with the =2
GR
ECS device as a potential means to accurately 250
quantify clay in Guata sands. In various Guata
wells, several runs of Platform Express integrated
Gamma ray, gAPI

wireline logging tools, including the ECS sonde, 200

systematically demonstrated poor linearity

between natural gamma ray and clay volume, Vclay,
150
from SpectroLith processing. They also showed
higher end points than are typical for sandstones,
both for the clean and shaly fraction (right).
100

19. The Waxman-Smits equation is described in reference 12.

20. The Sen-Goode-Sibbit saturation model is commonly
applied in freshwater shaly sand environments. For more 50
information: Sen PN, Goode PA and Sibbit A: Electrical
Conduction in Clay-Bearing Sandstones at Low and High
Salinities, Journal of Applied Physics 63, no. 10 (May 15,
1988): 48324840. 0
0 20 40 60 80 100
Clay volume, %
> Crossplot of gamma ray versus clay volume, as determined by SpectroLith
processing. The regression line highlights the lack of linearity between
measured gamma ray (GR) and clay volume, Vclay. The end points of the
regression line, at the intersection with the Vclay = 0 and Vclay = 1 axes, are
unusually high for a sandstone environment. These measurements were
later used in DecisionXpress processing.

Summer 2005 29
 Irreducible Water
Clay
Water
Quartz/Feldspar/Mica
Oil
Carbonate
Calcite
Pyrite Coates-Timur
Irreducible Water
Permeability
Pyrite
Grain Density from Water
Washout SpectroLith Processing 10,000 mD 0.1
Quartz
SDR Permeability Oil
Bit Size 2.5 g/cm 3 3
10,000 mD 0.1 Bound Water
6 in. 16 Grain Density from Core ELANPlus Fluid Analysis T2 Distribution
3
Klinkenberg- Illite
Caliper 2.5 g/cm 3 50 % 0 0 29
Corrected Core
6 in. 16 Gamma Ray Clay Volume from XRD Permeability Core Porosity ELANPlus Analysis T 2 LM

MD, ft 0 gAPI 300 0 % 100 10,000 mD 0.1 50 % 100 100 % 0 0.3 ms 5,000

X,450

X,500

X,550

X,600

> SpectroLith processing of a Guata log. Caliper measurements in the depth track show the hole is in good condition. Track 1, scaled
from 0 to 300 gAPI, shows high gamma ray values in the logged interval. SpectroLith clay volumes (gray) shown in Track 2 match core
measurements (blue circles); the SpectroLith grain density (red curve) is more reliable than the low core-density measurements (open
circles) from unconsolidated core samples. Permeability estimates in Track 3 match those measured in cores (blue circles). Computed
porosity (Track 4) also matches porosity measurements in cores (blue circles). Track 5 displays lithology and porosity from ELANPlus
volumetric analysis. NMR data in Track 6 show a well-developed free-uid signal in the highly permeable Guata sandstones.

Simultaneously, eld results from DecisionXpress
processing correlated strongly with results
obtained using a traditional ELANPlus advanced
multimineral log analysis technique in the Cara-
cas, Venezuela, computing center.
PDVSA found these results encouraging, and
the company decided to verify them by acquir-
ing a core and performing a series of X-ray
diffraction (XRD) measurements for compari-
son with the clay volume determined by
SpectroLith processing.
This comparison reveals a good correspon-
dence between total clay content determined by
XRD analysis and clay content determined by
SpectroLith analysis, even though the analyzed
core plugs were obtained in the cleaner,
more porous reservoir intervals (above).
Agreement between estimated porosity and esti-
mated permeability and core data was also
excellent. Some discrepancy remains between
grain density estimated from SpectroLith analy-
sis and grain density measured in core samples,
with core grain density typically lower than the
density of pure quartz. This discrepancy might
have resulted from the difficulty of accurately

30 Oileld Review
 Clay

Quartz/ Feldspar/ Mica

Mudcake Clay-Bound Water Carbonate

Washout Capillary-Bound Water Pyrite

Cable Tension Hydrocarbon Anhydrite

10,000 lbf 0
Moved Water Siderite
Bit Size
Free Water Coal
6 in. 16
Water
Calibrated Caliper Water Moved Hydrocarbon Salt
6 in. 16 Hydrocarbon
Hydrocarbon Data Quality Data Quality
Net Reservoir k-
DecisionXpress Flow Profile Permeability Total Porosity Total Porosity
Net Pay L PKSR
Mineralogy 0 1 10,000 mD 0.1 50 % 0 100 % 0

X,450

X,500

X,550

> DecisionXpress processing of the Guata log. To better evaluate the quick-look results obtainable using DecisionXpress
processing, the same interval of the previous log from Guata (previous page) was reprocessed using the DecisionXpress
system; this display is a default presentation. Caliper data in the depth track conrm hole quality was good except for rugosity
from X,465 to X,470 ft. This thin interval of data, masked with gray, is not reliable enough for automated interpretation. Net
reservoir and net pay ags are also shown in the depth track. Mineralogy from DecisionXpress processing appears in Track 1.
Track 2 shows the estimated production prole, derived from the relative permeability results shown in Track 3. Porosity and
uid information in Tracks 4 and 5 complete the evaluation. Mineralogy, shown in Track 5, is interpreted from ECS data using
DecisionXpress processing. Summarized in quality-control Track 6 are lithology (L), porosity (P), permeability (K), saturation (S)
and relative permeability (R); green indicates a favorable condition, yellow represents a moderately favorable condition and
red means an unfavorable condition. The quick-look DecisionXpress analysis agrees with core data and with the more time-
consuming ELANPlus analysis.

measuring grain density from essentially uncon-
solidated core plugs.
XRD analysis also showed the predominant
clay mineral to be kaolinite, often greater than
70% of the total clay, with the rest of the clay
minerals illite and a small fraction of chlorite. In
such conditions, a low mean CEC of the clays
would be expected; a value of 0.2 meq/g was
used in the DecisionXpress processing (above).
This quick-look result is remarkably similar to
the complete ELANPlus evaluation, including
the permeability estimate obtained from the
mineralogical version of the k- equation. The
rugose hole region from X,465 to X,470 ft is prop-
erly agged, and the main sandstone reservoirs
are properly diagnosed as being at or near irre-
ducible water saturation. This was conrmed by
production results, with the well coming on pro-
duction at a rate of 1,200 bbl/d [191 m3/d] of
uid, with a water cut of less than 20%.
DecisionXpress technology is now an integral
part of formation evaluation in the Guata eld,
ensuring that reliable interpretation results
from a remote region of Venezuela are available
wherever decisions need to be made, minutes
after the logs are acquired.

Summer 2005 31

W
R
Block 15/25b
O
N
N o r t h S e a
DENMARK
UK 0 km 200
0 mi 200
> OILEXCO license in the Outer Moray Firth
basin. A thin oil column, identied in 1990, led
the company to reevaluate the potential of
Block 15/25b.
Making Timely Decisions in the UK
In recent years, the UK government has encour-
aged owners of undeveloped offshore UK
discoveries to either develop or relinquish
acreage containing these discoveries. Many
unappraised prospects have therefore been
returned by the former owners and offered as
new licenses. The availability of these licenses
has attracted a number of operators new to the
North Sea who saw economic potential in some
of these relinquished blocks. One such operator
was OILEXCO, a Calgary-based company that is
currently developing Block 15/25b in the Outer
Moray Firth basin.
A thin oil column, discovered in 1990,
attracted the attention of OILEXCO (above).
After reprocessing seismic data and mapping
possible stratigraphic traps, the company initi-
ated a multiwell drilling program.21 To assist in
understanding results from logging, the com-
pany used the DecisionXpress system. Of the
32
Y
A
three wells logged with the DecisionXpress
system, Well 15/25b-8 proved to be the one that
justified further activity in the area, known as
the Brenda accumulation. That well was drilled
on an anomalous amplitude variation with offset
(AVO) elastic impedance response, and encoun-
tered a hydrocarbon column within the Forties
sandstone of nearly 50 ft [15 m] (next page).22
Timely petrophysical analyses using the
DecisionXpress system facilitated the rapid deci-
sion making required to sidetrack or to run
casing and test the wells. In addition, prompt
analyses provided OILEXCO with important eco-
nomic information to keep remote partners and
other investors fully apprised of reservoir capac-
ity and likely producibility. The resultant
appraisal work conrmed that the Brenda accu-
mulation may be one of the largest discoveries in
UK waters in recent years, and development
drilling using high-angle and horizontal wells
will commence in January 2006.
21. For more about exploration for stratigraphic traps by
OILEXCO: Durham LS: Subtle Traps Become New Prey,
AAPG Explorer 25, no. 8 (August 2004): 14.
22. Amplitude variation with offset (AVO) refers to a differ-
ence in seismic reection amplitude with a change in
distance between shotpoint and receiver. AVO
responses indicate differences in lithology and uid con-
tent in rocks above and below the reector.
23. For additional examples: Poulin M, Hidore J, Sutiyono S,
Herron M, Herron S, Seleznev N, Grau J, Horkowitz J,
Alden M and Chabernaud T: Deepwater Core Compari-
son with Answers from a Real-Time Petrophysical
Evaluation, paper SPE 90134, presented at the SPE
Annual Technical Conference and Exhibition, Houston,
September 2629, 2004.
Rasmus JC, Horkowitz JP, Chabernaud T, Graham P,
Summers M and Wise D: A New Formation Evaluation
Technique for the Lower Tertiary in South Texas
Predicting Production in Low Permeability, Fine-Grained
Sandstones, paper SPE 90690, presented at the SPE
International Petroleum Conference, Puebla, Mexico,
November 79, 2004.
Oileld Review
 Clay

Quartz/Feldspar/Mica

Mudcake Clay-Bound Water Carbonate

Washout Capillary-Bound Water Pyrite

Cable Tension Hydrocarbon Anhydrite

10,000 lbf 0
Moved Water Siderite
Bit Size
Free Water Coal
6 in. 16
Water
Calibrated Caliper DecisionXpress Water Moved Hydrocarbon Salt
6 in. 16 Mineralogy Hydrocarbon
Hydrocarbon Data Quality Data Quality
Net Reservoir Environmentally k-
Corrected Gamma Ray Flow Profile Permeability Total Porosity Total Porosity
Net Pay L P KSR
0 gAPI 200 0 1 10,000 mD 0.1 50 % 0 100 % 0

X,100

X,150

> Petrophysical analysis of Well 15/25b-8. The net pay ags in the depth track of this DecisionXpress display reveal
nearly 50 ft of oil pay near X,150 ft. Track 1 presents the gamma ray curve and lithology determined by the
DecisionXpress system. Track 3 shows hydrocarbon and water along with intrinsic permeability. Fluid saturations
and porosity are shown in Track 4. Detailed mineralogy, presented in Track 5, is determined using the ECS sonde and
DecisionXpress processing. As in all DecisionXpress presentations, a gray mask indicates that results are outside
tolerance specications.

Interpretation in Real Time
The DecisionXpress system has been applied
successfully to a wide range of siliciclastic reser-
voirs.23 This real-time interpretation is not fully
applicable to carbonate reservoirs, largely due to
the lack of a universally accepted, robust satura-
tion-evaluation scheme. On the other hand,
the lithology and matrix property components of
the system can bring signicant improvement to
carbonate evaluations, and they will be imple-
mented in the future.
By limiting the number of parameters
selected by the log interpreter, automated
interpretation minimizes interpretation bias.
As with any effort to automate tasks ordinarily
performed by people, automated interpreta-
tions must be carefully compared with other
data to ensure valid results. Comparisons of log
data to core and production data are critical for
operators to use this technology, but with time,
well-to-well comparisons should prove ade-
quate to validate interpretations.
The algorithms in the DecisionXpress system
yield fast and reliable petrophysical interpreta-
tion. By understanding how much hydrocarbon
is present, and whether it can be produced
economically, operating companies can better
plan formation-pressure and sampling jobs,
mechanical and percussion sidewall-coring jobs
and formation-testing operations, or elect to run
casing, drill ahead, or sidetrack. In addition,
rapid petrophysical analysis supports long-term
decision making, such as development of com-
pletion strategies, stimulation programs and
other operations. JS/GMG

Summer 2005 33
 New Fibers for Hydraulic Fracturing

Efficient proppant transport is essential to successful hydraulic fracturing. While conventional

fracturing fluids rely on high fluid viscosities, a new approach employs synthetic fibers that provide
excellent proppant transport at low fluid viscosities. This technology, which has improved well
productivity in several fields, gives engineers more flexibility when designing fracturing treatments.

Craig H. Bivins
Bivins Energy Corporation
Dallas, Texas, USA

Curtis Boney
Chris Fredd
John Lassek
Phil Sullivan
Sugar Land, Texas

John Engels
Houston, Texas

Eugene O. Fielder
Devon Energy
Oklahoma City, Oklahoma, USA

Tim Gorham
Chevron
McKittrick, California, USA
Fibers have been used in industry since
antiquity. Ancient Egyptians used straw and
Tobias Judd
horsehair to reinforce mud bricks. Early Chinese
Mexico City, Mexico
and Japanese houses show evidence of straw
Alfredo E. Sanchez Mogollon mats to provide structural support.1 However,
Reynosa, Mexico until synthetic fibers became commercially
available during the early 20th Century,
Lloyd Tabor commercial applications were limited by the
Oklahoma City, Oklahoma properties of naturally occurring fibers.
Today, a wide variety of manufactured fibers
Ariel Valenzuela Muoz is available, mostly made from polymers, metals,
PEMEX Exploracin y Produccin
glass or carbon. These fibers have properties that
Reynosa, Mexico
are revolutionizing many industries, particularly
civil engineering, medicine, apparel and
Dean Willberg
Moscow, Russia transportation. The oil and gas industry,
especially the pumping-services sector, is also
For help in preparation of this article, thanks to benefiting from new fibrous materials.
Ali Mazen, Sugar Land, Texas, USA; and Dharmesh Prasad, In the 1960s, engineers began adding nylon
Moscow, Russia.
CemNET, FiberFRAC, FracCADE, POD and PropNET are fibers to well cements for structural
marks of Schlumberger. reinforcement. The fibers transmit localized
stresses more evenly throughout the cement
matrix; as a result, the hardened cement is less
susceptible to stress cracking and shattering
during perforating.2
In the 1990s, Schlumberger introduced
CemNET advanced fiber cement, which
employed glass fibers to prevent lost circulation.3
As a CemNET cement slurry flows across a lost-
circulation zone during primary cementing, the
fibers form a bridging network and limit slurry
loss from the annulus to the formation. This
technology helps operators fill the annulus
completely with cement, improving zonal
isolation and avoiding remedial cementing.4
Fibers are also used to prevent proppant
flowback, a serious problem associated with
hydraulic fracturing.5 If proppant flows out of a
hydraulic fracture into the casing, well
productivity declines, and damage to casing,
control valves and wellhead equipment may

34 Oilfield Review
result. Pumped together with proppant in a
fracturing uid, the bers form a network that
stabilizes the proppant pack (below).6 To
maintain proppant-pack integrity, the fibers
must be sufciently stable to remain in place
during the productive life of the well. Today,
three PropNET hydraulic fracturing proppant-
pack additives, made from glass or polymer
bers, address a wide variety of well conditions.
Recently, Schlumberger researchers dis-
covered that, in addition to stabilizing a
proppant pack, fibers could enhance the
proppant-transport capabilities of fracturing
uids. Development of this concept in both the
laboratory and the field resulted in the
introduction of FiberFRAC ber-based fracturing
uid technology.
This article describes how bers improve
proppant transport, discusses the practical
advantages of using bers, and shows how the
technology can be employed to improve
hydraulic fracturing treatments. Case histories
from California, east Texas, the US midcontinent,
and northern Mexico illustrate the benets of
the ber technology.
How Fibers Prevent Proppant Settling
Hydraulic fracturing treatments comprise two
basic uid stages. The rst stage, or pad, does
not contain proppant, and is pumped through
casing perforations at a rate and pressure
sufficient to break down the formation and
create a fracture. The second stage, or proppant
slurry, transports proppant through the perfo-
rations into the open fracture. When pumping
ceases, the fracture closes onto the proppant.
> PropNET hydraulic fracturing bers in a proppant
pack. During well cleanup and production, the ber
network prevents proppant ow out of the fracture.
Summer 2005
Casing
Fracturing fluid
Proppant
Casing
Proppant
> Effects of proppant-settling velocities. High settling velocities cause
proppant to concentrate at the bottom of a fracture before it closes (top).
Low settling velocities promote complete and uniform distribution of
proppant throughout the fracture (bottom).
During injection and fracture closure, the rate of
proppant settling greatly inuences the nal
propped-fracture geometry.7
High settling velocities cause proppant to
concentrate at the bottom of a fracture before it
closes. In extreme cases, the proppant particles
form clusters that prevent further uid injection.
1. Li VC: Large Volume, High-Performance Applications of
Fibers in Civil Engineering, Journal of Applied Polymer
Science 83, no. 2 (2002): 660686.
2. Carter LG, Slagle KA and Smith DK: Stress Capabilities
Improved by Resilient Cement, API Drilling and
Production Practices. Washington, DC: American
Petroleum Institute (1968): 2937.
3. Low N, Daccord G and Bedel J-P: Designing Fibered
Cement Slurries for Lost Circulation Applications: Case
Histories, paper SPE 84617, presented at the SPE
Annual Technical Conference and Exhibition, Denver,
October 58, 2003.
4. Abbas R, Jarouj H, Dole S, Junaidi EH, El-Hassan H,
Francis L, Hornsby L, McCraith S, Shuttleworth N,
van der Plas K, Messier E, Munk T, Ndland N,
Svendsen RK, Therond E and Taoutaou S: A Safety Net
for Controlling Lost Circulation, Oileld Review 15, no. 4
(Winter 2003/2004): 2027.
Pay zone
Fracturing fluid
Pay zone
In either situation, proppant does not completely
ll the fracture, and well productivity suffers. By
contrast, low settling velocities promote a
complete and uniform distribution of proppant
throughout the fracture, and provide the greatest
potential for reservoir stimulation and
productivity improvement (above).
5. Proppants are sized particles mixed with fracturing uid
to hold fractures open after a hydraulic fracturing
treatment. In addition to naturally occurring sand grains,
man-made or specially engineered proppants, such as
resin-coated sand or high-strength ceramic materials
like sintered bauxite, may also be used. Proppant
materials are carefully sorted for size and sphericity to
provide an efcient conduit for uid ow from the
reservoir to the wellbore.
6. Card RJ, Howard PR and Fraud J-P: A Novel
Technology to Control Proppant Backproduction,
paper SPE 31007, SPE Production and Facilities 110,
no. 4 (November 1995): 271276.
Armstrong K, Card R, Navarrette R, Nelson E, Nimerick K,
Samuelson M, Collins J, Dumont G, Priaro M,
Wasylycia N and Slusher G: Advanced Fracturing Fluids
Improve Well Economics, Oileld Review 7, no. 3
(Autumn 1995): 3451.
7. Constein VG: Fracturing Fluid and Proppant
Characterization, in Economides MJ and Nolte KG (eds):
Reservoir Stimulation, 2nd edition. Englewood Cliffs,
New Jersey, USA: Prentice Hall (1989): 5-15-23.
35
 Casing

Fracturing fluid

Pay zone

Propped fracture

Casing
Fracturing fluid

Propped fracture Pay zone

> Effect of viscosity on fracture geometry and proppant placement. Excessive uid viscosity causes fracture growth
outside of the pay zone, promoting proppant settling and suboptimal production (top). Fluids with lower viscosities
reduce the fracture height (bottom). Adding bers helps to suspend the proppant until the fracture closes.

Proppant transport in conventional fracturing
uids is governed by a complex combination of
parameters, including particle size and density,
fracture dimensions and base-uid rheological
properties. Fluid viscosity is particularly
important because it provides resistance to
gravitational settling, and helps transport the
proppant along a fracture. Several studies have
investigated proppant-settling rates versus uid
viscosity.8 One overarching fluid-viscosity
guideline has emerged from these studies and
field experience: for conventional fracturing
uids, the minimum uid viscosity to ensure
adequate proppant transport is about 100 cP at
100 s1 shear rate.9
Fracturing-uid viscosity also affects fracture
geometry. As uid viscosity increases, fracture
width increases. Unfortunately, the bottomhole
treating pressure also increases, potentially
causing excessive vertical fracture-height
growth. If the fracture grows beyond the pay zone
into nonproductive or water-producing intervals,
the overall efciency of the fracturing treatment
suffers (above).
Therefore, engineers must design fracturing
uids that transport proppant efciently, while
keeping fractures within the pay zones. In many
areas, achieving both goals is difcult, which
sometimes forces engineers to make compromises
that lead to suboptimal results. Fortunately, bers
offer a solution to this dilemma.
Adding bers to a uid-particle suspension
dramatically alters particle-settling behavior.
When bers are not present, settling generally
proceeds according to Stokes law.10 The speed at
which particles fall through a uid is directly
proportional to particle size and density, and
inversely proportional to fluid viscosity. As
sedimentation proceeds, a distinct boundary
forms between the particle bed and the uid
lying above.
With bers present, Stokes law no longer
applies. The bers interfere with the particles,
physically hindering their downward journey. As
sedimentation proceeds, a distinct particle-
liquid boundary does not develop; instead, the
ber-particle mixture slowly compresses, leaving
little uid behind. This type of behavior is called
Kynch sedimentation (next page, top).11
The practical benet of Kynch sedimentation
is that uid viscosity plays a much smaller role in
determining the particle-settling velocity.
Experiments show that, at a given base-uid
viscosity, fibers reduce the proppant-settling
velocity by more than one order of magnitude.
Equivalently, at a given settling rate, the
required base-uid viscosity also decreases by
about one order of magnitude (next page,
bottom). Indeed, in terms of particle
sedimentation, one may think of the bers as
providing virtual fluid viscosity. Use of
FiberFRAC technology reduces the importance
of base-fluid viscosity as a settling-rate
determinant, giving engineers much wider
latitude when designing a fracturing treatment.

36 Oileld Review
Optimizing Fibers for Proppant Transport Conventional FiberFRAC
Appropriate bers for proppant transport must proppant slurry slurry
have the correct combination of length, diameter,
exibility and temperature stability. The bers
must be easy to disperse in a proppant slurry, and
able to pass through pumping equipment,
tubulars and perforations without breaking or
bridging. The bers cannot separate from the
proppant during placement. After placement, the
bers must be stable until the fracture closes.
However, unlike PropNET fibers, proppant-
transport bers should dissolve after fracture
closure to maximize proppant-pack conductivity.
Schlumberger scientists experimented with
many types of bers before nding products that
satised all these requirements. They chose two
fibers that cover two reservoir-temperature
ranges: 150F to 250F [66C to 121C] and
250F to 400F [204C]. During the development
of FiberFRAC technology, two laboratory
Stokes law sedimentation Kynch
evaluations were particularly important: the slot sedimentation
test and the proppant-pack conductivity test.
A slot test is a dynamic laboratory technique
to evaluate proppant transport. The test
apparatus has a transparent slot that simulates
an 8-ft long, 1-ft high and 516-in. wide [2.44-m by > Stokes law and Kynch sedimentation. In conventional fracturing uids,
30.4-cm by 0.47-cm] fracture. Proppant slurry proppant particles settle according to Stokes law, forming a distinct
ows through an orice simulating a perforation boundary between the proppant bed and the uid lying above (left). Proppant
and then passes across the slot, permitting slurries containing bers display Kynch sedimentation (right). As sedimentation
proceeds, a distinct particle-liquid boundary does not develop; instead, the
observation and measurement of proppant- ber-proppant mixture (inset) slowly compresses, leaving little uid behind.
transport efficiency. One test compared the
performance of conventional and FiberFRAC
slurries, each using the same base fluid.
Proppant of 20/40-mesh size was added at a

8. Novotny EJ: Proppant Transport, paper SPE 6813,

presented at the SPE Annual Fall Technical Conference
and Exhibition, Denver, October 912, 1977.
Roodhart LP: Proppant Settling in Non-Newtonian No fiber
Initial settling rate, mm/min

Fracturing Fluid, paper SPE/DOE 13905, presented at the 100

SPE/DOE Low Permeability Gas Reservoirs Symposium,
Denver, March 912, 1985.
Acharya A: Particle Transport in Viscous and
Viscoelastic Fracturing Fluids, paper SPE 13179, SPE 10
Production Engineering 1, no. 2 (March 1984): 104110.
.
9. Shear rate, , is the velocity gradient measured across
the diameter of a uid-ow channel, such as a pipe,
annulus or other shape. In most oileld viscometers, the
shear rate is the velocity difference between a rotating 1
sleeve and a cylinder (or bob) mounted concentrically
inside the sleeve. When uid is present in the annulus
between the bob and sleeve, the bob experiences 1 10 100
torque as the sleeve rotates. This force is called shear Fluid viscosity, cP
stress, . Viscosity, , is the shear-stress to shear-rate
.
ratio, = /. The viscosity of many uids varies with
shear rate. Therefore, viscosity specications must 0.00 % fiber by weight of fluid
include the shear rate. In this article, the shear rate for
0.75 % fiber by weight of fluid
all viscosities is 100 s1.
(2gr 2)(d1d2) 1.00 % fiber by weight of fluid
10. According to Stokes law, V = , where V is
9 1.50 % fiber by weight of fluid
the particle-fall velocity (cm/s), g is the acceleration of 2.00 % fiber by weight of fluid
gravity (cm/s2), r is the equivalent particle radius (cm),
d1 is the particle density (g/cm3), d2 is the uid density > Proppant-settling velocities with and without bers. When bers are
(g/cm3) and is the uid viscosity (dyne-s/cm2). present, the particle-settling rate in a proppant slurry is more than one order
11. Tiller FM: Revision of Kynch Sedimentation Theory, of magnitude slower than that observed in a uid without bers (blue). At a
American Institute of Chemical Engineers Journal 27, given settling rate, the required base-uid viscosity also falls by about one
no. 5 (1981): 823828. order of magnitude.

Summer 2005 37
 concentration of 2 lbm [0.9 kg] to each gallonUS
50-gallon [3.8 L] of fracturing uid (2 ppa).12 The pump
mixing tanks rates for the conventional and FiberFRAC
slurries were 26.6 and 17.4 gal/min [101 and
66.1 L/min], respectively. Despite the higher
1-ft x 8-ft x 516-in. slot pump rate, proppant in the conventional uid
and 516-in. perforation
separated and fell to the bottom of the slot. The
FiberFRAC slurry was stable, bers were evenly
dispersed, and all proppant remained in
suspension during the test (left). In addition, the
FiberFRAC Slurry
bers did not break or bridge while passing
through the perforation.
Proppant-pack conductivity is a direct function
of the void space between the proppant particles;
therefore, it would be best if the FiberFRAC bers
disappeared. Unlike PropNET bers, which
must maintain a rigid network, FiberFRAC bers
are no longer needed after proppant-slurry
2 ppa 20/40-mesh proppant at 17.4 gal/min placement and fracture closure. For this reason,
Schlumberger scientists chose polymers that
Conventional Proppant Slurry slowly dissolve (below left). Laboratory testing
conrmed that, once the bers dissolved,
the resulting proppant-pack conductivity was
essentially identical to that obtained from the
same uid without bers (next page, top).

Wellsite Delivery
Proper execution of a hydraulic fracturing
treatment requires smooth and stable mixing of
2 ppa 20/40-mesh proppant at 26.6 gal/min all fluid components at the correct concen-
trations. Fibers have a high length-to-diameter
> Slot tests comparing the performance of FiberFRAC ber-based fracturing ratio; consequently, adding and dispersing them in
uid and conventional proppant slurries. A schematic diagram of the slot a proppant slurry may be a challenge. Fortunately,
apparatus (top) shows the uid path. The photographs show proppant slurries
owing across the slot. Both tests used the same base uid with 2 ppa of this problem was addressed previously, during the
20/40-mesh ceramic proppant. The FiberFRAC slurry is stable (middle), while development of PropNET technology.
proppant in the conventional slurry (bottom) falls to the bottom of the slot. The POD programmable optimum density
blender is equipped with a special feeder for
adding bers to fracturing uids (next page,
bottom). This feeder comprises a hopper into
which the bers are poured, and an auger that
100 blends the bers into the proppant slurry at a
90 steady rate. During large treatments, a conveyer
80 belt transports bers into the feeder.
Fiber decomposition time, days

Low-temperature fiber
70 The physical and chemical properties of
60
FiberFRAC bers differ from those of PropNET
fibers. Therefore, before pumping actual
50
treatments, mixing tests were necessary to verify
40
High-temperature fiber the suitability of the existing feeder on a POD
30
blender. Minor auger recalibration compensated
20 for the different bulk density and ow properties
10 of dry FiberFRAC bers. Actual eld treatments
0 could then proceed.
150 175 200 225 250 275 300 325 350 375 400
Temperature, F 12. Proppant concentrations are commonly expressed in
pounds of proppant added, or ppa. One ppa means
> Decomposition rates of two types of FiberFRAC bers. The low-temperature that one pound of proppant is added to each gallon of
fracturing uid. It must not be confused with the more
ber (blue) is used at temperatures between 150 and 250F. The high- common pounds per gallon, or lbm/gal. During
temperature ber (red) is used at temperatures between and 250 and 400F. hydraulic fracturing treatments, ppa better reects eld
Both bers decompose within days to weeks. practice. There is no recognized metric equivalent to ppa.

38 Oileld Review
 100
90 Press frame
200F 275F
Retained permeability, %

80
68% 66% 69% 68%
70
60 Heated platens
50
40
30
20 Force actuator Core Core
10
0 Conductivity
No 30 lbm/1,000 gal No 30 lbm/1,000 gal cell
fiber FiberFRAC fiber fiber FiberFRAC fiber
Guar concentration: 18 lbm/1,000 gal Guar concentration: 20 lbm/1,000 gal

Pressure transducer
> Effect of FiberFRAC bers on proppant-pack conductivity. The conductivity Backpressure
test (right) involves placing a proppant slurry between two cores, and regulator
Fluid leakoff
inserting the proppant sandwich in a conductivity cell. The cores are Core
heated and compressed in the cell to simulate downhole conditions, and
uid from the proppant slurry leaks off through the cores. After leakoff, brine
is pumped through the proppant pack, and the pressure drop is recorded.
FiberFRAC bers have a minimal effect on the proppant-pack conductivity
(left). In these experiments, the proppant was 20/40-mesh bauxite, and the
closure stress was 5,000 psi [35 Mpa].

Proppant pack

Fiber hopper

POD blender

> Wellsite delivery. A hopper above the POD programmable optimum density blender feeds bers into the fracturing uid.

Summer 2005 39
Stimulating the Monterey Formation
in California Well A Well B
Chevron initiated the first applications of
Total depth, ft 2,860 3,181
FiberFRAC technology at the Lost Hills eld in
Bottomhole 125 125
California, USA.13 The oil-producing Monterey static temperature,F
formation is composed of diatomite with a high Hole size, in. 8.75 8.75
Perforation
porosity45 to 65%and a relatively low density, shots/ft 4 4
permeability of 1 to 7 mD. The reservoir pressure
Perforated intervals, ft
ranges from 500 to 1,400 psi [3.5 to 9.8 MPa], at
an average bottomhole temperature of 125F Stage 1 2,610 to 2,620 2,835 to 2,845
[51.7C]. The pay zone is 800 to 1,200 ft [152 to Stage 2 2,340 to 2,350 2,630 to 2,640, 2,660 to 2,670
366 m] thick at an average depth of about 1,800 ft Stage 3 2,120 to 2,130 2,420 to 2,430, 2,460 to 2,470
Stage 4 1,750 to 1,760
[549 m]. The stress barriers lying above and
below the pay zone are weak, and may not
Total sand, lbm 1,252,000 1,230,000
contain a hydraulic fracture. Total fluid, gal 327,000 334,000
The most interesting feature of this diatomite
is its softness. The Youngs modulus is extremely
Treatment overview for Lost Hills A and B wells, California, USA.
low50,000 to 300,000 psi [345 to 2,070 MPa]
about one order of magnitude less than that of
hard sandstones.14
These rock properties make conventional
fracturing fluids difficult to use. The high 70
viscosities of crosslinked fracturing uids create
1,000 lbm proppant/ft of interval

excessive near-wellbore fracture width. Because

60
of the weak stress barriers, there is little control
of fracture height. Previous treatments with a
bbl/day

50
conventional 350-cP [0.35-Pa.s] borate-
crosslinked guar uid system usually resulted in
40
short, wide fractures that extended beyond the
pay zone. This ballooning of the fracture also
caused signicant proppant-owback problems. 30
Average oil production,

Schlumberger engineers used the FracCADE

fracturing design and evaluation software to
determine the uid properties required to create
longer, narrower fractures that would be conned
within the pay zone. The computer simulations
predicted that the uid viscosity should be less
than 100 cP [0.1 Pa.s]. In light of the proppant-
transport guidelines described earlier, Chevron
stimulated the next group of wells with
FiberFRAC technology. After preliminary
laboratory testing, a 33-cP [0.03-Pa.s] linear
(uncrosslinked) guar solution was chosen as the
base uid. The FracCADE simulator also showed
that the FiberFRAC treatments would require
less proppant, because the fractures would be
narrower and conned within the pay zone.
The Lost Hills hydraulic fracturing program
benefited from surface tiltmeter fracture
mapping performed by Chevron while fracturing
with conventional uids.15 The company also
installed tiltmeters during seven FiberFRAC
treatments, giving engineers a rare opportunity
to compare fractures created by both types
of uids.
20 Fiber, 1,700 lbm proppant/ft of interval
Crosslinked fluid, 2,100 lbm proppant/ft of interval
10 Crosslinked fluid, 2,500 lbm proppant/ft of interval
0
30-d average 60-d average 90-d average
> Average production from Lost Hills wells treated by conventional and
FiberFRAC uids. The production rates are normalized to the amount of
proppant placed in the fractures. After 90 days, the FiberFRAC wells were
more efcient producers than their conventional counterparts.
The seven treatments involved multiple zones showed that the average fracture length was
in two wells (top). Well A was treated in four 182 ft [55.5 m], compared to 145 ft [44.2 m] for
stages, Well B in three. Bridge plugs provided the conventional treatments. As predicted by
zonal isolation between the stages. The simulations, much less proppant was required to
perforations were oriented to induce fracture achieve the longer fractures: 1,700 lbm of
growth in the preferred direction.16 proppant per foot of interval [2,530 kg per meter],
The FiberFRAC treatments were pumped compared with 2,100 to 2,500 lbm per foot [3,130
successfully, without fiber bridging at the to 3,730 kg per meter] for the conventional jobs.
perforations or difficulties placing proppant After 90 days, the production rates from wells
inside fractures. Analysis of the tiltmeter data treated with conventional and ber-laden uids

40 Oileld Review
 Well 1 Well 2
Dolomite Dolomite
Limestone Limestone
Sand Sand
Shale Shale
> Lithology of Wells 1 and 2 in the Lower Cotton
Valley formation. The logs conrm a nearly
identical sand series in the two wells.
were the same; however, the FiberFRAC wells
required 30% less proppant. On a typical Lost
Hills well, this is equivalent to obtaining the
same production rates with 720,000 lbm
[327,000 kg] less proppant (previous page,
bottom). In addition, because of narrower near-
wellbore fracture width, little or no proppant
owback occurred during production.
Improving Production in the Lower Cotton
Valley Formation
The Lower Cotton Valley formation of east Texas
and northern Louisiana, USA, consists of thin,
laminated, gas-producing sandstones between
shale layers. The permeability ranges between
0.001 and 0.05 mD. Gas-producing formations
with such low permeabilities are often classied
as tight gas. Reservoir depths range from 10,000
to 14,000 ft [3,048 to 4,270 m], and the
bottomhole temperatures vary between 200 and
Summer 2005
340F [93 and 182C]. The Youngs modulus of
the sandstone is 5,000,000 psi [34,470 MPa],
more than one order of magnitude greater than
the Lost Hills diatomite mentioned earlier.
For this geological situation, the main
fracturing design challenges include the need for
long fractures, limited fracture-height growth,
good proppant coverage over the entire fracture
surface and minimal proppant-pack damage. In
addition, the fracturing uid must be stable at
high temperatures.
Many operators have performed conventional
massive hydraulic fracturing treatments in this
formation, using crosslinked polymer fluids.
Unfortunately, these treatments often create
large fractures that extend into nonproductive
zones, and require large volumes of polymer
and proppant, reducing the economic viability
of stimulation.
On the other end of the viscosity spectrum,
fracturing with slick-waterwater plus a
friction reduceris a popular stimulation
method for this formation.17 The viscosity of slick-
water is about 1 cP. Exerting sufcient force to
initiate and propagate a fracture during the pad
stage, and to transport proppant, requires high
pump rates, usually greater than 50 bbl/min
[7.9 m3/min]. The proppant concentration in the
uid is low, usually less than 2 ppa. The proppant
size is usually small40/70 meshto minimize
the Stokes law settling rate. This method is
much less expensive than massive hydraulic
fracturing, and has greatly expanded the number
of wells that can be economically stimulated.
However, further investigations revealed that,
when normalized for reservoir and producing-
system conditions, wells stimulated in this
manner were less productive than those
treated conventionally.18 Despite the high pump
rates, small proppant sizes and low proppant
concentrations, proppant tends to settle
relatively close to the wellbore, limiting the
effective fracture length.
Neither conventional nor slick-water treat-
ments can address all of the stimulation
challenges posed by the Cotton Valley formation.
This area requires a low-viscosity fracturing uid
that transports proppant efciently. Following
the success of ber-laden fracturing uids in
California, FiberFRAC technology was proposed
for the Cotton Valley formation.19 The first
application occurred in the Okio eld, operated
by Bivins Operating Company. The producing
reservoir depth was 13,000 to 14,000 ft [3,960 to
4,270 m], and the bottomhole temperature was
260F [127C].
Petrophysical analysis revealed that, because
of the high Youngs modulus, the minimum uid
viscosity needed to achieve sufcient hydraulic
fracture width for ber placement was 50 cP
[0.05 Pa.s]. To prevent excessive fracture-height
growth, the upper uid-viscosity limit was 150 cP
[0.15 Pa.s]. At a uid temperature of 260F, a
delayed borate-crosslinked guar uid satised
these requirements. The polymer concentration
was 18 lbm/1,000 galUS [2.2 kg/m3]. Without
fibers present, 30 to 35 lbm of guar per
1,000 galUS [3.6 to 4.2 kg/m3] would be necessary
to achieve adequate proppant transport. In
addition to reducing the fluid cost, using
less polymer improves the proppant-pack
conductivity and increases well productivity.20
The FiberFRAC treatments were completed
on a group of offset wells with similar lithological
characteristics (left). The permeabilities and
zone heights for both wells, collectively
expressed as kh, were essentially the same: 0.30
13. Vasudevan S, Willberg DM, Wise JA, Gorham TL,
Dacar RC, Sullivan PF, Boney CL and Mueller F:
Field Test of a Novel Low Viscosity Fracturing Fluid
in the Lost Hills Field, California, paper SPE 68854,
presented at the SPE Western Regional Meeting,
Bakerseld, California, USA, March 2630, 2001.
14. Youngs modulus, E, is an elastic constant that indicates
how a material deforms when subjected to stress.
Resistance of a material to deformation increases with
the value of E.
VanVlack LH: A Textbook of Materials Technology.
Reading, Massachusetts, USA: Addison Wesley (1973):
1112.
15. Surface tiltmeter fracture mapping directly measures
fracture orientation, fracture volume, geometry and
approximate location. Tiltmeters measure miniscule
changes in surface tilt at many points around the well
during a hydraulic fracturing treatment. They operate
like a carpenters level. The tiltmeter data are analyzed
to determine the fracture parameters that would
produce the observed deformation eld.
Barree RD, Fisher MK and Woodroof RA: A Practical
Guide to Hydraulic Fracture Diagnostic Technologies,
paper SPE 77442, presented at the SPE Annual
Technical Conference and Exhibition, San Antonio,
Texas, USA, September 29October 2, 2002.
16. Acock A, Heitmann N, Hoover S, Zia Malik B,
Pitoni E, Riddles C and Solares JR: Screenless
Methods to Control Sand, Oileld Review 15, no. 1
(Spring 2003): 4243.
17. Mayerhofer MJ and Meehan DN: Waterfracs: Results
from 50 Cotton Valley Wells, paper SPE 49104,
presented at the SPE Annual Technical Conference and
Exhibition, New Orleans, September 2730, 1998.
18. England KW, Poe BD and Conger JG: Comprehensive
Evaluation of Fractured Gas Wells Utilizing Production
Data, paper SPE 60285, presented at the SPE Rocky
Mountain Regional/Low Permeability Reservoir
Symposium and Exhibition, Denver, March 1215, 2000.
19. Engels JN, Martinez E, Fredd CN, Boney CL and
Holms BA: A Mechanical Methodology of Improved
Proppant Transport in Low-Viscosity Fluids: Application
of a Fiber-Assisted Transport Technique in East Texas,
paper SPE 91434, presented at the SPE Eastern
Regional Meeting, Charleston, West Virginia, USA,
September 1517, 2004.
20. Nimerick KH, Temple HL and Card RJ: New pH-
Buffered Low-Polymer Borate-Crosslinked Fluids,
paper SPE 35638, presented at the SPE Gas Technology
Conference, Calgary, April 28May 1, 1996.
41
and 0.28 mD-ft.21 Well 1 was treated with ber- Initial production from Well 1 was 3.1 million To date, more than 120 FiberFRAC
laden uid; slick-water was used on Well 2. For ft3/d [87,800 m3/d], while that from Well 2 was treatments have been performed in east Texas
Well 1, engineers pumped the 18-lbm/1,000 galUS 0.70 million ft3/d [19,800 m3/d]. After 90 days of and Louisiana. More than 22 million lbm
borate-crosslinked guar system, and the production, Well 1 produced at an average rate of [10 million kg] of proppant have been placed,
proppant slurries contained 1 to 6 ppa 20/40- 1.9 million ft3/d [53,800 m3/d], while the average using more than 7 million gallons [26,500 m3] of
mesh sand. On average, 390,000 lbm [176,900 kg] production rate from Well 2 was 0.66 million ft3/d fiber-laden fluid. Bottomhole temperatures
of proppant were placed during each treatment. [18,700 m3/d]. During this period, the cumulative varied between 197 and 339F [93 and 182C],
By contrast, the proppant slurries during the gas production from Well 1 was seven times with the majority pumped at about 265F
slick-water treatments contained 0.25 to 4 ppa greater than that of Well 2 (below). [129C]. The largest ber-laden treatment to
40/70-mesh sand, and the average amount of date placed 850,000 lbm [385,000 kg]
proppant placed was 200,000 lbm [90,700 kg]. of proppant.

Improving Gas Production in the

10.0 Barnett Shale
Well 1fiber-assisted technology The Barnett shale is among the fastest growing
Well 2slick-water fracture onshore gas elds in the United States. Devon
Gas rate, million ft3/day

Energy operates 550,000 acres [222,530 ha] in

1.0
0.1
0 10
> Gas production from Wells 1 and 2 in the Lower Cotton Valley formation. The well stimulated with
ber-laden fracturing uid (dark blue) produced considerably more gas during the rst 130 days of
production than the well stimulated with slick water (gray).
100
Production normalized to interval height, million ft3/ft
the Fort Worth basin in northern Texas. The 200-
to 600-ft [60-to 180-m] thick Barnett shale is
even tighter than the Lower Cotton Valley
sandstone. The permeability is 0.0001 mD, and
the Youngs modulus ranges from 2 to 3 million
psi [13,790 to 20,680 MPa]. The average depth
and bottomhole temperature are 8,000 ft and
200F [2,440 m and 93C], respectively.
20 30 40 50 60 70 80 90 100 110 120 130
Time, days The stimulation history of the Barnett shale is
similar to that of the Lower Cotton Valley
sandstone. Operators initially performed massive
hydraulic fracturing treatments, involving uids
with high polymer concentrations and nearly 1.5
million pounds [680,400 kg] of proppant. The
treatment costs were high, and the resulting
production was often insufficient to justify

commercial development of the Barnett shale.

To reduce treatment costs, many operators
80
switched to slick-water fracturing fluids.22
FiberFRAC well production data Smaller sand treatments, usually involving about
60 200,000 lbm of proppant and increased liquid
volume, were promising and greatly increased
40
the number of wells that could be stimulated
economically. However, certain areas did not
Offset well production data
fulll their production potential. The estimated
20 ultimate recovery (EUR) of the problem areas
was generally greater than 1,000 million ft3 [28
0 million m3].23 In these tight formations, fracture
0 30 60 90 120 150 180 length largely determines well productivity.
Time, days
Because slick-water has limited transport
capability, the effective fracture lengths were not
> Production after stimulation in the Barnett shale. Wells stimulated with
FiberFRAC uid (blue) outperformed those stimulated with slick-water (red).
sufficient to achieve the desired well
When normalized to the interval height, the FiberFRAC wells were twice as productivity.
productive. To obtain more effective proppant distribution
and improve well productivity Devon Energy used
ber-laden fracturing uids. The fracturing uid
21. kh is the product of the formation permeability (k) and 23. Estimated ultimate recovery (EUR) is dened as the
the zone height (h), expressed in mD-ft. quantity of hydrocarbon that is estimated to be was a borate-crosslinked guar system with a
22. Brister BS and Lammons L: Waterfracs Prove potentially recoverable from an accumulation, plus the polymer concentration of 18 lbm/1,000 galUS. The
Successful in Some Texas Basins, Oil & Gas Journal 98, quantity already produced.
proppant concentration varied from 0.5 to 3 ppa.
no. 12 (March 20, 2000): 7476. 24. Tapia NGE, Ruiz JM, Ramisa LR, Mengual JF and
Cern AS: Well Construction and Field Development in The treatment placed 540,000 lbm [245,000 kg] of
Mexico, Oileld Review 15, no. 4 (Winter 2003/2004): 20/40-mesh sand. Following the treatment,
4653.

42 Oileld Review
 Well 1
Neutron Porosity
Depth,
Gamma Ray ft Caliper
2,850
2,900
> Openhole logs of offset wells in Arcabuz-Culebra eld. Well 1 (left), treated conventionally, had nearly identical petrophysical characteristics when
compared with Well 2 (right), treated with FiberFRAC uid.
production was monitored and compared with
that of conventional slick-water treatments
(previous page, bottom). During the rst 80 days of
production, the wells treated with FiberFRAC
technology delivered an additional 25 million ft3
[708,000 m3] of gas compared with the offset wells.
Improving Gas Recovery in Northern Mexico
The Arcabuz-Culebra eld, operated by PEMEX
Exploracin y Produccin in northern Mexico, is
part of the larger Burgos basin that extends into
south Texas, USA. The pay zone is the Wilcox
formation, a gas-producing sandstone that is
frequently associated with water-producing
intervals.24 Today, the basin produces around
1,000 million ft3/d, and PEMEX is working to
double this production rate.
The Youngs modulus of the Wilcox sandstone
is 4 to 4.5 million psi [27,580 to 31,030 MPa], and
the permeability ranges from 0.001 to 0.05 mD.
Most wells are drilled to depths of 9,500 to
9,800 ft [2,895 to 2,987 m], where the bottomhole
temperature is about 250F. These formation
characteristics require a low-viscosity fracturing
fluid to minimize fracture entry into the
water-producing layers. Because of the low
formation permeability, long fractures are
needed to maximize well productivity. Therefore,
Schlumberger proposed FiberFRAC technology
as a solution.
Summer 2005
Gamma Ray
The rst treatment took place near offset
wells with similar kh values. The kh value of
Well 1 was 86.3 mD-ft, while that of Well 2 was
94.7 mD-ft. (above). Before stimulation, Well 1
produced 300,000 ft3/d [8,500 m3/d] of gas, while
Well 2 did not produce at all.
Well 1 was treated conventionally with a
30-lbm/1,000 galUS [3.6-kg/m3] borate-crosslinked
guar system, placing 200,000 lbm of ceramic
proppant into the fracture. The FiberFRAC
treatment in Well 2 employed a less
viscous, 20-lbm/1,000 galUS [2.4-kg/m3] borate-
crosslinked guar uid and, like Well 1, 200,000 lbm
of proppant were placed at concentrations
ranging from 1 to 8 ppa. The pump rate for
both treatments was 30 bbl/min [4.8 m3/min].
Following the treatments, the gas production rate
from Well 2 was more than ve times greater than
that of Well 1. In addition, the water production
rate from Well 2 was only half that of Well 1. In
light of these results, more fracturing treatments
with ber-laden uids are planned.
Well 2
Neutron Porosity
Depth,
ft Caliper
2,900
2,950
The Future of Fiber-Assisted Transport
The eld application of FiberFRAC technology is
still in its infancy; however, initial results
confirm the promise demonstrated during
laboratory development. Low-viscosity, fiber-
laden fracturing uids appear to be particularly
appropriate for stimulating formations that
require careful fracture-height control.
Fibers also provide mechanical support for
the proppant as it travels down a fracture,
maximizing the effective fracture length. This
attribute is particularly valuable when
stimulating low-permeability gas formations.
Fiber-laden fluids are probably not
economically justifiable for stimulating wells
with an EUR below about 1,000 million ft3 of gas.
Under these circumstances, slick-water appears
to be the more appropriate fracturing uid.
An increasing number of elds could benet
from FiberFRAC technology. Candidate selection
is presently under way in the US and Canadian
Rocky Mountains, and in Russia. Further use of
FiberFRAC technology will more clearly
dene the window of well conditions in which it
is appropriate. EBN
43
 Using Casing to Drill Directional Wells

Drilling with large-diameter tubulars eliminates the need to run conventional

drillpipe, which then must be pulled to install permanent casing. It can mitigate lost
circulation, enhance well control and reduce nonproductive rig time, and also
decrease the risk of unintentional sidetracks or stuck pipe. Fewer trips into and out
of a well, improved wellsite safety, increased efciency and lower cost have led to
an expanding range of applications that now includes casing directional drilling.

Kyle R. Fontenot The use of casing for drilling is an emerging
ConocoPhillips technology that can reduce well-construction
Puerto La Cruz, Venezuela costs, improve operational efficiency and
safety, and minimize environmental impact.
Bill Lesso Fundamentally simple in principle, this drilling
Houston, Texas, USA technique uses the large-diameter tubulars that
will be permanently installed in a wellbore in
R. D. (Bob) Strickler
place of conventional drillpipe. The economic
ConocoPhillips
Houston, Texas demands of complex geologic settings, smaller
reservoirs with limited recoverable reserves, and
Tommy M. Warren the need to optimize development and
Tesco Corporation exploitation of mature fields make drilling
Houston, Texas operations with casing increasingly attractive to
operating companies.
For help in preparation of this article, thanks to Lee Conn, A conventional rotary drill bit or a special
M-I SWACO, Houston; and Mike Williams, Sugar Land, Texas.
ABC (Analysis Behind Casing), PowerDrive, PowerDrive drilling shoe can be attached to the end of a
Xceed and PowerDrive Xtra are marks of Schlumberger. casing string to drill vertical wells. For additional
Casing Drilling, Casing Drive System and Drill Lock exibility and for those applications that require
Assembly (DLA) are marks of Tesco Corporation.
DrillShoe is a mark of Weatherford. directional control, a retrievable bottomhole
EZ Case is a mark of Hughes Christensen, a division of assembly (BHA) for drilling can be deployed,
Baker Hughes, Inc. locked in place and then retrieved by wireline
cable. Running and retrieving this BHA through
casing eliminate tripping of drillpipe into and
out of a wellbore and provide added protection
for the advanced systems used in downhole
measurements and directional-drilling applications.
Minimizing the number of pipe trips during
drilling operations reduces incidents of hole
collapse from swabbing and surging, decreases
the chance of an unintentional sidetrack and
minimizes wear inside previously installed
surface or intermediate casing strings. After
reaching total depth (TD), the casing is already
in place, eliminating the need to pull drillpipe
and then run permanent casing.
This reduction in pipe handling improves
wellsite safety and allows drillers to use standard-
size rigs or smaller rigs built specically to drill
with casing. New compact rigs for drilling
operations with casing require lower horsepower,
use less fuel, produce fewer emissions, operate
from smaller surface locations, and can be moved
more quickly and easily than larger conventional
rigs (next page).
Compared with traditional drilling operations,
casing while drilling minimizes rig downtime
resulting from unexpected occurrences, such as
stuck pipe or loss of well control from an inux,
or kick, of formation uid. Anecdotal evidence
indicates that drilling with larger diameter
tubular connections reduces lost circulation by
mechanically plastering cuttings and drilled
solids into the borehole wall.

44 Oileld Review
 It is possible that this smearing effect
builds an impermeable ltercake or creates a
solid surface nish that may allow weak, low-
pressure and depleted intervals to be drilled
without signicant loss of drilling uid.
Casing strings have longer joints than
standard drillpipe, which means that drillers
make about 25% fewer connections. Another
benet is less time spent circulating uid or
backreaming to maintain hole stability while
making pipe connections. In addition to
improving drilling efficiency, both of these
advantages further reduce overall cost and
environmental impact.
Drilling operations with casing eliminate
several steps in the conventional well-
construction process and provide other critical
advantages, including better uid circulation
and removal of formation cuttings for more
effective hole cleaning. As operators gain
experience in an area, drilling penetration rates
with casing usually improve, ultimately matching
or surpassing penetration rates previously
achieved with drillpipe when comparing days per
1,000 ft [305 m] or feet per day.
Analysis of wells drilled to date with casing
indicates that this technique can reduce
nonproductive rig time by as much as 50% and
cut drilling time by a nominal 10 to 35% per well
in some applications. About one-third of this
reduction results from decreased tripping of
pipe; the remainder comes from avoiding
unexpected drilling problems and from
eliminating the time required to install casing in
a separate operation.
This faster, simpler and more efficient
process translates into fewer drilling surprises
and lower costs. Advances in tools, equipment
and procedures are expanding the use of this
technology for drilling soft and hard formations
both onshore and offshore, and most recently for
casing directional drilling.
We rst review the use of casing for drilling,
including ongoing inll development activity in
south Texas, and then discuss how > Casing while drilling and casing directional drilling. During the past ve years, ConocoPhillips and
simultaneously drilling and casing a well helps Tesco Corporation drilled extensively with casingmore than 1,050,000 ft [320,040 m]in south
reduce borehole problems. The results from Texas, recently expanding applications to include directional operations and compact purpose-built
rigs, such as the one shown here. This technique improved drilling efciency and effectively
recent testing of directional operations with eliminated lost circulation in about 110 wells. These results and similar experience in other areas
casing demonstrate how rotary steerable system indicate that casing can be used to avoid lost circulation and drill through pressure-depleted zones
(RSS) technology improves drilling efciency in mature elds that are difcult to drill using conventional drillpipe, onshore and offshore.
compared with steerable downhole motors,
especially for smaller borehole sizes.

Summer 2005 45
A Fundamental Change in Well Construction
Both positive-displacement motor (PDM) and
RSS technologies utilize drillpipe. This specially
designed, thick-wall pipe is run to the bottom of
a borehole and pulled back out, perhaps several
times while drilling a well, and again to install
and cement a permanent string of casing during
a separate operation distinct from the rest of the
drilling process.
Introduction of the downhole PDM in the
1960s facilitated drilling without full-string pipe
rotation. These systems use mud owing through
a turbine or a rotor-stator power section to
generate torque downhole. Steerable motors
with a xed bend angle, or bent housing, allow
simultaneous control of borehole azimuth and
inclination angle, which subsequently resulted in
better directional control and routine
construction of high-angle wellbores, horizontal
borehole sections in the 1980s and, eventually,
extended-reach wells in the 1990s.
In the late 1990s, rotary steerable systems
helped operators set new records in extended-
reach drilling (ERD). This technology, including
the Schlumberger PowerDrive and PowerDrive
Xtra rotary steerable systems, and the
PowerDrive Xceed rotary steerable system for
harsh, rugged environments, facilitates
directional control and steering of the bit while
continuously rotating the entire drillstring.
Roller-cone or xed-cutter bits on the end of
rotating drillpipe have monopolized the drilling
of oil and gas wells for a century. However, new
concepts and design improvements in rotary rigs
and drill bits have been the norm since these
tools were introduced in the early 1900s. As a
result, penetration rates and bit life have
improved dramatically during this period.1
Using casing to drill oil and gas wells
represents a fundamental change in the process of
constructing a wellbore. Casing while drilling
provides the same hole-making capability as
drillpipe operations, with better removal of drilled
cuttings and improved hole-cleaning performance.
The casing used for drilling can be a partial liner or
a full string (right). From its earliest applications
wellbore tubulars also have been used for
slimhole drilling at various times since the 1950s.
In the 1960s, Brown Oil Tools, now Baker Oil
Tools, patented a relatively advanced system for
drilling with casing that included retrievable
pilot bits, underreamers to enlarge hole size,
and downhole motors. However, low penetration
rates compared with conventional rotary
drilling restricted the commercial application
of this system.2
Research and development continued at a
slow pace until the late 1980s, when economic
and market conditions stimulated renewed
interest in drilling with conventional tubing,
coiled tubing and other slimhole techniques. At
about the same time, Amoco, now BP,
documented success drilling and coring with
mining equipment and tubulars. In the 1990s,
operators began using liners to drill from
normally pressured formations into pressure-
depleted intervals.
This approach avoided problems, such as hole
instability and enlargement, lost circulation and
well control, which plagued conventional drilling
operations. Mobil, now ExxonMobil, used partial
liners to drill from higher pressure transition zones
into the extremely depleted limestone reservoirs of
Conventional Retrievable BHA
Drilling for Liner Drilling
Drillpipe
Surface
casing
Liner
Intermediate hanger
casing
Liner
PDC bit
the Arun gas eld in North Sumatra, Indonesia.3
Amoco also employed this technique to drill wells
in the Norwegian North Sea Valhall eld.4
In 2001, BP and Tesco reported success using
casing to drill surface and production casing
intervals for 15 gas wells in the Wamsutter area
of Wyoming, USA. These wells ranged in depth
from 8,200 to 9,500 ft [2,499 to 2,896 m].5 At
about the same time, Shell Exploration and
Production Company dramatically improved
drilling performance in south Texas by drilling
underbalanced with casing, realizing a cost
reduction of about 30%.6
To date, operators have drilled more than
2,000 wellbore sections using casing. More than
1,020 of these intervals have involved vertical
drilling with casing and nonretrievable bits,
about 620 were drilled using partial liners, more
than 400 used a retrievable BHA for vertical
drilling, and about 12 used a retrievable BHA
for directional drilling. All of these early
applications helped casing while drilling evolve
from a new technology with unproven reliability
to a practical solution that can reduce costs,
increase drilling efficiency and minimize
nonproductive rig time.
Nonretrievable Drilling Retrievable Drilling
BHA for Casing BHA for Casing
Production
Underreamer casing
Drill
PDC shoe
pilot bit

until the recent surge in activity, using casing for
drilling has shown signicant potential compared
with conventional drilling.
In the 1920s, the Russian oil industry
reported the development of retractable bits for
use in drilling operations with casing. In the
1930s, operators in the continental USA used
production tubing to drill openhole, or barefoot,
completions. The tubing string and the at-
blade, or shtail, bit used for drilling remained in
the well after production ow began. Permanent
> Simultaneous drilling and casing with partial liners or full casing strings. Drilling operations
traditionally involved roller-cone or xed-cutter bits on the end of rotating drillpipe (left). As
alternatives to this standard approach, operators and service companies developed and tested
various systems for drilling with liners and casing. Drilling with a partial liner uses enough pipe to
case the open hole and omits the upper part of the casing string (middle left). Conventional drillpipe
conveys the bottomhole assembly (BHA) to target depth and carries the main drilling loads. A liner
hanger or packer connects the drillstring with the liner. The BHA can be retrieved only when the hole
is nished. If a BHA failure occurs, the entire drillstring and liner must be pulled. Liner hanger
position within the previous casing string limits the maximum drilling depth. A full string of casing
with a nonretrievable drillable bit (middle right) or a retrievable drilling BHA (right) provides
additional functionality and exibility. The retrievable BHA can be deployed and retrieved with
smaller jointed pipe, coiled tubing or wireline cable without tripping casing into and out of a well.

46 Oileld Review
 A retrievable system allows the bit and BHA
7-in. casing to be deployed initially and replaced without
to surface tripping casing into and out of the hole.8 This
option is the only practical choice for directional
Drill Lock wells because of the need to recover expensive
Assembly (DLA) BHA components, such as downhole motors,
rotary steerable systems or measurements-while-
drilling (MWD) and logging-while-drilling (LWD)
8 joints of tools. A wireline-retrievable system facilitates
7 5/8-in. casing replacement of equipment that fails before
reaching TD, and allows quick, cost-effective
Tandem internal
casing stabilizers
access to log, evaluate and test formations.
Several service providers are committed to
developing tools, techniques and equipment for
Casing shoe drilling with casing. Tesco, for example, offers
Casing Drilling services, comprising purpose-built
rigs, surface equipment and downhole tools for
Drill collar spacer onshore applications.
joint, or float sub
To facilitate the use of casing for drilling, Tesco
designed robust, reliable surface equipment and
6 1/8-in. to 8 7/8-in. downhole systems that efciently and effectively
underreamer
attach to and release from casing. A wireline-
conveyed drilling assembly is typically suspended
in a prole nipple near the bottom of a casing
string. The Tesco Casing Drilling system uses a
Tandem external
pilot-hole stabilizers Drill Lock Assembly (DLA) to anchor and seal
the BHA inside casing (left).9

1. Besson A, Burr B, Dillard S, Drake E, Ivie B, Ivie C,

Smith R and Watson G: On the Cutting Edge,
Oileld Review 12, no. 3 (Autumn 2000): 3657.
6 1/8-in. PDC pilot bit 2. Hahn D, Van Gestel W, Frhlich N and Stewart G:
Simultaneous Drill and Case TechnologyCase Histo-
ries, Status and Options for Further Development, paper
> Casing vertical drilling. The retrievable bottomhole assembly (BHA) for vertical drilling
IADC/SPE 59126, presented at the IADC/SPE Drilling
includes a small bit that drills a guide, or pilot, hole (left). An underreamer with expandable Conference, New Orleans, February 2325, 2000.
and retractable cutter pads expands this initial borehole to accept the full diameter of the 3. Sinor LA, Tybero P, Eide O and Wenande BC: Rotary
casing being used. Stabilizers between the pilot bit and the underreamer maintain borehole Liner Drilling for Depleted Reservoirs, paper IADC/
inclination. Upper stabilizers located inside the casing reduce BHA vibrations and protect SPE 39399, presented at the IADC/SPE Drilling Conference,
the Drill Lock Assembly (DLA), which provides an axial and torsional connection to the Dallas, March 36, 1998.
casing (right). The Tesco DLA seals against the casing to direct drilling uid through the bit. 4. Tessari RM and Madell G: Casing Drilling
A Revolutionary Approach to Reducing Well Costs,
It also allows uid to bypass the BHA during wireline deployment and retrieval. A positive- paper SPE/IADC 52789, presented at the SPE/IADC
displacement motor (PDM) or a rotary steerable system (RSS), heavyweight drill collars, Drilling Conference, Amsterdam, March 911, 1999.
measurements-while-drilling (MWD) systems or logging-while-drilling (LWD) tools, not 5. Shepard SF, Reiley RH and Warren TM: Casing Drilling:
shown here, can be included. The DLA is run on wireline and landed in a prole nipple An Emerging Technology, paper IADC/SPE 67731,
near the bottom of the casing. The BHA is positioned in the last casing joint so that all presented at the SPE/IADC Drilling Conference,
components below the tandem stabilizer extend into the open hole below casing. Amsterdam, February 27March 1, 2001.
6. Gordon D, Billa R, Weissman M and Hou F:
Underbalanced Drilling with Casing Evolution in the
South Texas Vicksburg, paper SPE 84173, presented at
the SPE Annual Technical Conference and Exhibition,
A New Approach Conventional rotary bits that remain in the Denver, October 58, 2003.
Some operators now view this technology as a wellbore after reaching TD have been used in 7. Hossain MM and Amro MM: Prospects of Casing While
potential solution in a variety of commercial some applications. The bit can remain on the Drilling and the Factors to Be Considered During Drilling
Operations in Arabian Region, paper IADC/SPE 87987,
applications, ranging from drilling entire casing and be cemented in place, or it can be presented at the IADC/SPE Asia Pacic Drilling
onshore wells to drilling just one or two hole released and dropped into the bottom of the hole Technology Conference and Exhibition, Kuala Lumpur,
September 1315, 2004.
sections in offshore wells that require multiple to allow logging. Drillable drill bits, such as the 8. Warren T, Tessari R and Houtchens B: Casing Drilling
casing strings.7 Drillers categorize the downhole Weatherford Type II or Type III DrillShoe or the with Retrievable Drilling Assemblies, paper OTC 16564,
presented at the Offshore Technology Conference,
systems that are used to drill with casing as Baker Hughes EZ Case, have external cutting Houston, May 36, 2004.
nonretrievable or retrievable. A nonretrievable, structures for drilling, but can be removed by 9. Warren TM, Angman P and Houtchens B: Casing Drilling
or xed, assembly can be used to drill with short milling. These specially designed casing shoes Application Design Considerations, paper IADC/SPE
59179, presented at the IADC/SPE Drilling Conference,
liners or full casing strings. allow drilling and completion of subsequent New Orleans, February 2325, 2000.
borehole sections. Shepard et al, reference 5.
Warren T, Houtchens B and Madell G: Directional
Drilling with Casing, paper SPE/IADC 79914, presented
at the SPE/IADC Drilling Conference, Amsterdam,
February 1921, 2003.

Summer 2005 47
 Each BHA component must pass through the pilot hole. A topdrive unit rotates the casing and connections. An internal spear assembly
casing string that is used for drilling, including applies torque to make tubular connections. provides a uid seal inside the pipe.
an underreamer, or hole enlarger, with The Tesco quick-connect Casing Drive Initially, drilling operations with casing were
retractable pads. A pilot bit initiates a small System, which is operated by the topdrive performed onshore in vertical wells to avoid the
hole, which then is enlarged by cutters on the hydraulic control system, speeds up pipe additional complexity of offshore operations. As a
expanded underreamer pads. Drillers commonly handling and prevents damage to casing threads result, casing vertical drilling advanced to a point
use a 618-in. or a 614-in. pilot bit and an by eliminating one cycle of making and breaking where it routinely rivaled the efficiency of
underreamer that expands to 878 in. when drilling connections at tubular joints (below).10 A slip operations with conventional drillpipe. Tesco
with 7-in. casing. The underreamer can be assembly grips either the exterior or interior of Corporation and ConocoPhillips have drilled more
located immediately above the bit outside the the casing, depending on pipe size, and attaches than 100 of these vertical wells in south Texas.
casing or above other BHA components in the the casing to the topdrive without threaded
A Proving Ground in South Texas
ConocoPhillips initiated an infill-drilling
6 5/8-in. connection program in 1997 to increase production and
to topdrive unit recovery from geopressured Wilcox sands in the
south Texas Lobo geologic trend. Operators
discovered natural gas in these low-permeability,
Hydraulic actuator
or tight, sands near the USA and Mexico border
in the 1960s, but limited well productivity, low
gas prices and inadequate pipeline capacity
made commercial development uneconomic.
From 1979 to the mid-1990s, US tax
10 ft [3 m]
incentives for tight-gas development, advances in
hydraulic fracture stimulation, new pipeline
construction and higher gas prices resulted in
the drilling of more than 1,000 wells. Since 1997,
Axial and ConocoPhillips has drilled another 900 wells,
torque grapple
ranging in depth from 7,500 to 13,000 ft [2,286 to
3,962 m], to recover additional gas reserves in
Packer cup this area.
Most of these wells were drilled in a single run
Casing stabbing
guide with conventional drillpipe and polycrystalline
> Surface equipment for casing while drilling. The Tesco Casing Drive System comprises a quick- diamond compact (PDC) fixed-cutter bits.
connect slip assembly that grips either the exterior (left) or interior (middle) of the casing, depending on Despite extensive experience in this mature area,
pipe size, and attaches the casing to the topdrive without threaded connections to prevent thread drilling efciency peaked in 2001 after about
damage. An internal spear assembly provides a uid seal inside the pipe. The Casing Drive System is 600 wells. Rig downtime represented less than
operated by a topdrive system that is suspended from the derrick block, so the entire topdrive rotary
mechanism is free to travel up and down (right). A topdrive differs radically from the more 10% of the total time to drill a Lobo well, so a new
conventional rig-oor rotary table and kelly method of turning the drillstring because it allows drilling approach was required to reduce well-
to be performed with three joints at a time instead of single joints of pipe. It also allows drillers to construction costs further.
quickly engage the rig pumps or the rotary drive while tripping pipe, which minimizes both the
In 2001, ConocoPhillips began reevaluating
frequency of stuck pipe and the cost per incident.
well-construction practices to increase drilling
efciency enough to make exploitation of smaller
Lobo reservoirs with less than 1,000 million ft3
2000 2001 [28.3 million m3] of recoverable gas economical.
This would allow development activity to
Stuck pipe
8% 4% 4% continue for several years in this highly faulted
8% 34% Well control 7% 37%
and compartmentalized geologic trend.
3%
3% Stuck pipe Lost circulation
Stuck pipe
Even though surface, intermediate and
3% production intervals could be drilled
Cementing
Lost circulation conventionally, downhole drilling problems and
Lost circulation Fluids
rig downtime near the TD of each casing section
Directional control
continued to impede performance. Lost
3%
38% Mechanical 39% 9% circulation, stuck drillpipe and inability to run
Lay down casing casing to TD were common in Lobo trend wells,
> Nonproductive, trouble time for conventionally drilled wells in the south Texas Lobo trend. Lost accounting for about 75% of trouble time in 2000
circulation and stuck pipe were found to be the major causes of trouble encountered while drilling and 2001 (left).
Lobo eld wells with conventional drillpipe. In 2000 and 2001, these two problems accounted for 72%
and 76%, respectively, of the trouble time. Well control and inability to successfully run casing to TD
also were signicant in these and other years.

48 Oileld Review
 compete with conventional drilling across the
entire Lobo trend. This second phase proved that
drilling with casing mitigates formation-related rig
downtime associated with conventional operations.
Downtime on the next 11 wells drilled with
casing consisted primarily of rig mechanical and
operational-related problems; there were
virtually no incidents of stuck pipe or lost
circulation. In addition, many of the mechanical
and operational problems were reduced or
eliminated. During the rst two phases of this
program, the performance of Tesco Casing
Drilling systems steadily improved, matching the
average daily penetration rate of conventional
operations by the fifth well and eventually
1,400
exceeding it (left).
The surface-casing sections of wells in the
1,200
Lobo program were drilled with 958-in. casing
Rate of penetration (ROP), ft/day

1,000 using an 812-in. PDC pilot bit and a 1214-in.

underreamer in a retrievable BHA.
800 Conventional
drilling average ROP
ConocoPhillips drilled this interval in one run for
600 all of the wells and encountered few problems
retrieving the BHA with wireline. Actual drilling,
400 or rotating, times with casing were slightly
200 higher than conventional operations using
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
drillpipe and a 1214-in. rotary drill bit.
0
These 500-ft [152-m] sections were
6,8

8,1
6,8

8,6
8,0

7,1

6,7

6,2

7,3

8,7

8,9

7,8

6,9

8,1

8,4

8,7

completeddrilled, cased and cementedin

80

26

20
00

78

23

05

25

20

18

97

11

73

03

50

75

Order of Lobo wells drilled using casing with total footage about the same time as conventionally drilled
> Improvements in the efciency of casing while drilling. Rate of penetration surface holes. Cement inside the 958-in. casing
improved dramatically during the course of initial vertical drilling operations with was drilled out with 7-in. casing using a 614-in.
casing in the south Texas Lobo trend. By the end of Phase 1, a ve-well pilot PDC bit and 812-in. underreamer congured to
program, the performance of casing while drilling matched that of conventional
operations with drillpipe. Well 7 included 705 ft [215 m] of hole drilled directionally
mill and clean out inside casing. After drilling
with casing and a downhole motor. Well 8 included a 902-ft [275-m] section drilled through cement inside the casing and then into a
directionally with drillpipe and heavyweight drilling collars. few feet of formation below the casing set point,
or shoe, this BHA was retrieved and replaced
with another for drilling 878-in. open hole.
In early wells, this second BHA drilled to a
depth where formations become harder, typically
During conventional drilling operations, Many well-control incidents and blowouts about 6,500 ft [1,981 m]. A third BHA drilled to
additional fluid, or mud, often had to be occur while tripping pipe. Using casing to drill the 7-in. casing point. In most cases, the bit and
circulated to recondition the borehole and helps avoid these unexpected, dangerous and underreamer had little wear at either point.
address problems such as lost circulation, potentially costly events. Drilling operations with After gaining more experience, ConocoPhillips
sloughing formations and hole collapse in casing minimize or eliminate pipe trips and leave began drilling this entire intermediate casing
pressure-depleted intervals. Gas influx at casing at the bottom of a borehole, the best section in a single run.
intermediate casing points or across productive position to circulate out an inux. This is an
zones, and stuck-pipe conditions while drilling or important advantage, especially as this 10. Warren T, Johns R and Zipse D: Improved Casing
Running Process, paper SPE/IADC 92579, presented
running casing also were problems. As a result, technique is applied in more applications under at the SPE/IADC Drilling Conference and Exhibition,
well-control incidents were a major concern. increasingly complex subsurface conditions. Amsterdam, February 2325, 2005.
11. Fontenot K, Highnote J, Warren T and Houtchens B:
ConocoPhillips identied casing while drilling as The rst phase of evaluating drilling operations Casing Drilling Activity Expands in South Texas, paper
a technology that might solve these problems and with casing involved a ve-well pilot program. SPE/IADC 79862, presented at the SPE/IADC Drilling
Conference, Amsterdam, February 1921, 2003.
improve drilling efciency.11 Beginning in late 2001 and continuing into 2002,
ConocoPhillips extended this program to
determine whether casing while drilling could

Summer 2005 49
 Drill Openhole Reaming Back
with Casing Logging to Bottom
7-in.
casing shoe
4 12-in. casing
Stabilizer
Wireline
logging tools
Bit-release
mechanism
> Procedure for logging after drilling with casing. A technique for running
openhole wireline logs for formation evaluation that proved effective in
the Lobo development program was to drill to TD with 412-in. casing and
release the bit (left). The next step was to ream back to the 7-in. casing
shoe, so openhole logs could be run through the 412-in. casing just as
they would if the well were drilled conventionally (middle). The 412-in.
casing was then reamed back to TD (right).
Borehole sections for production casing in valve was set at the bottom of the casing. This
some of the rst wells of Phase 2 were drilled valve allowed cement to be pumped into the
with conventional drillpipe until procedures for borehole annulus, but prevented it from
drilling with 412-in. casing were established. The backowing, or U-tubing, into the casing. The
production sections of subsequent wells were casing then was reamed back to TD and
drilled with a 614-in. PDC bit attached to the end cemented in place.
of casing by a mechanical releasing device. This For wells that did not require openhole
device also functioned as a near-bit stabilizer, a logging, a slickline-conveyed oat valve was set
spacer joint, a crossover from casing connections in order to cement the casing in place through
to bit connections and a reaming shoe after the the bit. Nondrillable pump-down oat valves are
bit was released (above). available for some pipe sizes, including 7-in.
After reaching TD in wells that needed to be casing, and Tesco also has developed drillable
logged for formation evaluation, the bit was pump-down oat equipment. These cementing
released by dropping a ball. The 412-in. casing improvements allow casing and wellhead surface
was backreamed and pulled up into the 7-in. connections to proceed without having to wait
casing to allow openhole wireline logging. After for cement to set, which further minimizes
logging, a slickline-conveyed cementing float nonproductive rig time.
50
The initial success of this technique
reinforced the belief that drilling operations
using casing can be performed without
premature failure of tubular connections. During
Phases 1 and 2, casing with buttress threads was
used to drill surface and intermediate hole
sections. A torque ring installed in each casing
connection provided a torque stop and increased
the torque capacity of the coupling.
Manufacturers also are developing new
casing connections that can handle higher
torque. A special coupling, designed by
Grant Prideco, was used for drilling operations
with 412-in. casing. ConocoPhillips now uses this
coupling with 7-in. casing to drill intermediate
hole sections. Surface-casing hole sections
continue to be drilled using 958-in. casing with
buttress threads and a torque ring.
Casing while drilling has successfully
minimized trouble time related to lost
circulation and stuck pipe. The retrievable BHA
has been extremely reliable during running and
resetting at depths up to 9,000 ft [2,743 m].
Concerns about inclination control have been
reduced through proper BHA design.
Two nearby offset wells in the Lobo program
illustrated the benets of drilling operations
with casing. These wells did not require logs and
were drilled within seven months of each other. A
conventional rig operating in this area for more
than four years drilled the rst well. The second
well was the fteenth and, at that time, the
fastest well drilled using casing and a Tesco
Casing Drilling rig. Excluding rig repair time on
both wells, the conventional well took 300 hours
from start to rig release; the well drilled with
casing took 247.5 hours, a 17.5% reduction in
drilling time (next page, top).
The penetration rate for conventional drilling
was slightly faster than for casing while drilling.
However, the well drilled with casing
experienced only slight lost returns, and drilling
was able to continue after uid losses stopped.
Total downtime resulting from lost circulation
was less than an hour. In contrast, the
conventional well was plagued by uid losses
from about 6,500 ft [1,981 m] to the intermediate
casing point at about 9,500 ft and required an
additional 53 hours to deal with four lost-
circulation events.
Drilling operations with casing included only
66 hours of nonproductive rig time at
intermediate and production casing points
compared with 113.5 hours for the conventional
well. Neither well encountered significant
problems during drilling operations, so this
Oileld Review
 0 conventional drillpipe. Implementing an
1,000 effective solution to these two problems could
Well 16 drilled using casing cut total drilling time for a 9,500-ft well to about
2,000
Offset well drilled using drillpipe 200 hours, a 33% reduction from the previous
3,000 300 hours.
Measured depth (MD), ft 4,000 In Phase 3 of this program, ConocoPhillips
mobilized three new Tesco Casing Drilling rigs
5,000
built specically to drill in the Lobo trend (bottom
Lost circulation
6,000 left). These compact units include a topdrive to
7,000
handle larger derrick loads and an automated
power catwalk system that transfers casing to the
8,000
rig oor. They also offer increased fuel efciency
9,000 and require a smaller surface pad, or footprint.
10,000 The small, mobile Casing Drilling rigs have a
0 100 200 300 depth rating of 15,000 ft [4,572 m] and were
Time, hr designed for optimal drilling operations with
> Drilling time versus depth for conventional drilling and casing while casing, but also can use conventional drillpipe.
drilling. The rate of penetration (ROP) for the conventional well (blue) was During the past ve years, ConocoPhillips has
slightly faster than for casing while drilling (red), but operations were
plagued by lost circulation from about 6,500 ft to the intermediate casing
drilled more than 350 intervals and about
point at about 9,500 ft. 1,050,000 ft [320,040 m] in 110 wells using
retrievable drilling systems for casing.
Collectively, experience in these wells conrmed
that casing while drilling could eliminate or
reduce lost circulation and other problems
associated with depleted zones.
Initially drillers expected lost circulation to
be a problem when using casing to drill because
of the increase in equivalent circulating density
Water tank (ECD). A higher ECD results from the smaller
Mud pumps
annular clearance between large casing and the
Drillers
cabin Mud tanks and borehole wall, which increases frictional
Hydraulic solids control
power units pressure losses. The exact mechanism during
Automated
casing while drilling that mitigates lost
power catwalk circulation is not clearly understood at this time,
but combined with a higher ECD, it allows lower
mud weights to be used, which may facilitate air
drilling and underbalanced drilling.
Electrical
generators During all three phases of this Lobo
development project and other applications of
Fuel and
lubrication casing while drilling, no signicant or serious lost-
storage circulation events have occurred. Even in areas
near conventionally drilled wells that previously
required multiple remedial cement plugs and
> A more compact drilling rig. Tesco Casing Drilling rigs were designed on additional unscheduled full-length liners to reach
standard oileld skids, so the entire rig can be moved in 12 loads rather than TD, there were fewer lost-circulation problems
the 23 loads required for conventional rigs. The most modern conventional and fewer incidents of stuck pipe.12
rigs used in the Lobo development area require about 33 truckloads to make
a move, with move time averaging about 2.2 days. The new rigs can be This ConocoPhillips work established the
moved with standard oileld winch trucks without the use of a crane. A rig reliability of a retrievable drilling BHA and
move requires 12 hours from release to start of the next well. hinted at potential future applications for casing
while drilling. Several operators are pursuing
applications for this technique in areas where
difference reected the relative efciency of Penetration rates also have improved with conventional drilling costs are high. In these
these two methods at the casing points. However, experience, reducing drilling time by another applications, improvements in operational
about 17 hours were lost waiting for cement to 30 hours. Tests are under way to understand the efficiency would provide an even greater
set on the well drilled with casing. As improved lower rate of penetration (ROP) with casing, economic impact.
pump-down oat devices for all casing sizes which should help drillers increase casing
became available, this cementing downtime also penetration rates to equal or exceed those of 12. Fontenot et al, reference 11.

was reduced.

Summer 2005 51
 Increasing emphasis on redeveloping mature casing demonstrated the viability of casing for casing, the PDM and bent housing are placed
offshore properties in which high-angle wells directional drilling, but also highlighted the above the underreamer and pilot bit to rotate
must traverse pressure-depleted zones offers an limitations of steerable motors (below). both. This configuration allows slide drilling
excellent opportunity to drill directionally with without rotating the entire string to make
casing and realize signicant cost savings. Steerable Downhole Motors directional corrections. As a result, the BHA
However, only about 34,000 ft [10,363 m] in Drilling with casing and steerable motors in test geometry for directional control with steerable
12 well intervals have been drilled directionally wells and actual eld operations identied three motors and casing differs from a conventional
using a steerable PDM or an RSS in a retrievable limitationsBHA geometry, motor performance BHA for drillpipe (bottom left).13
BHA. These operations with 7-in. and 958-in. and operational practices. In a retrievable BHA In addition, drilling systems for casing
directional drilling must pass through the casing,
so the entire BHA and the PDM are smaller
Commercial Directional Wells Drilled with Casing
relative to the hole size. This limits the motor
Well Casing Initial Distance Maximum Build rate, Type of Type of
size, in. depth, ft drilled, ft inclination, degrees degrees/100 ft application BHA bend angle. The bent housing contact pad often
does not touch the borehole wall. Instead, a pilot-
1 9 5/8 339 2,993 4 2 Avoid collision PDM
hole stabilizer is incorporated below the
2 9 5/8 370 3,468 4 2 Avoid collision PDM
underreamer cutters to provide directional
3 7 6,000 705 8 1.5 Build angle and hold PDM
control and ensure a smooth borehole trajectory.
4 9 5/8 393 2,247 40 3 Build angle and hold PDM
Smaller motors and components also
5 9 5/8 393 3,172 17 1.5 Build angle and hold PDM increase BHA flexibility, so maintaining
6 9 5/8 492 1,968 16 1.5 Build angle and hold PDM directional control is more difcult. The entire
7 7 2,115 4,418 16 Drill tangent PDM assembly is tilted at a greater angle in the
8 9 5/8 633 2,739 17 2 Vertical and build angle PDM borehole and has a tendency to build inclination
9 7 4,434 3,427 15 2.5 S profile PDM angle, which makes dropping borehole angle
10 7 1,278 4,672 29 2.5 S profile RSS more difcult. Adding an expandable stabilizer
11 9 5/8 8,987 1,118 80 1.5 build Build angle PDM or an underreamer with noncutting stablilizer
1.5 turn and turn pads above the motor reduces rotating build
12 7 5,007 2,843 25 3 S profile RSS and rates and provides the capability of dropping
PDM
inclination angle by sliding, but this makes the
> Commercial directional wells drilled with casing. In its rst commercial application, casing BHA more complex.
directional drilling was used to drill surface holes to 3,332 ft [1,016 m] and 3,838 ft [1,170 m] with Another inefciency arises when PDM torque
958-in. casing for two offshore wellsWells 1 and 2, respectively. The most extensive commercial
reaches higher levels and the circulating
drilling using casing was conducted in Mexico, where 958-in. casing was used to kick off and build
inclination for intermediate hole sections in three wellsWells 4, 5 and 6drilled from a central pressure increases, causing the drillstring to
surface pad onshore. elongate. Because the bit is on bottom and the
casing cannot move downward, both the weight
on bit (WOB) and the required rotational motor
torque increase, further exacerbating the
increase in circulating pressure.14
Steerable Motor Assembly for Drillpipe PDC bit This effect is cyclic and causes motors to slow
Nonmagnetic drill collar MWD system Stabilizer Steerable PDM down and stop, or stall. The problem worsens
with casing, which tends to lengthen more under
internal pressure than does conventional
drillpipe. For a given internal pressure increase,
the additional WOB for 7-in. casing is about six
times greater than for 312-in. drillpipe with the
Steerable Motor Assembly for Casing PDC bit same size motor.
Underreamer
Nonmagnetic drill collar MWD system Steerable PDM In deeper wells and under conditions of high
borehole friction, increased WOB may be difcult
to detect at the surface. As a result, a PDM may
> Directional drilling geometry and control points. In a conventional directional BHA for drillpipe, stall before drillers can take corrective action.
three distinct pointsthe bit, a stabilizer pad on the motor bend and a stabilizer above the motor The consequence is that smaller, less powerful
dene the geometry for building inclination angle (top). The upper two points are noncutting, so the motors that are required for casing while drilling
geometry and stiffness of the BHA force the bit to cut along a circular path. In casing directional may have to be operated at less than optimal
drilling, three points also determine the build rate for a steerable motor, but the points are not as
torque and pressure to compensate for abrupt
dened and are more difcult to modify (bottom). The lower point is still the bit, but the second point
is not located at the motor bend. A smaller motor relative to the hole size must be used to pass WOB changes.
through the casing in a retrievable assembly. As a result, the motor bend often does not contact the The primary issue with smaller motors is a
borehole wall. Instead, a rotating, noncutting stabilizer below the underreamer cutter pads functions relative lack of power compared with larger
as the second control point. Directional control may be affected because the bit is farther away from
versions. Selection of the most appropriate motor
the upper control point.
for directional drilling is critical, particularly for

52 Oileld Review
7-in. and smaller casing sizes. Low-speed motors Vertical Plot Horizontal Plot
with more torque output in response to increased 0
pressure are easier to operate. A bit with less
400
aggressive cutting structures that do not cut as
deeply into the formation also improves motor 1,000
performance. All of these factors, however,

North-South departure, ft
300
reduce drilling efciency and penetration rates.
For casing larger than 958 in., motor power 2,000
considerations are less critical because larger
200
motors relative to hole size can be used. In some
cases, it may be advantageous to use motors
3,000
specifically designed for casing directional

True vertical depth (TVD), ft

100
drilling that provide high torque at relatively
low pumping pressure.
Recovering from motor stalls and 4,000
0
reorientating a BHA require less time with casing -300 -200 -100 0
because casing is stiffer than drillpipe. The East-West departure, ft
casing does not twist as much between the 5,000
surface and a PDM, so there is no need to
reciprocate casing to relax this stored torque.
The WOB is allowed to drill off without lowering
6,000
the casing. The BHA then is picked up slightly
and rotated to the desired orientation. If a motor
stalls, the pump rate is reduced and the string is
picked up to restart the motor, usually without 7,000

having to readjust the bend angle of the motor.

If borehole friction causes the casing to hang
up, manually or automatically rocking, or 8,000
0 1,000 2,000
rotating, the entire string forward and reverse,
Horizontal displacement, ft
without changing the BHA orientation helps
control abrupt changes in WOB when sliding.
This allows the motor to run more consistently
and improves drilling performance without
affecting directional control.15
PDM limitations and the potential benets of
using rotary steerable technology were evident in
south Texas drilling operations with casing.
ConocoPhillips drilled two wells in the Lobo
trend using a retrievable BHA with a PDM for
vertical inclination control. Two other Lobo wells > Lobo Well 83 vertical and horizontal trajectory plots. To avoid a surface obstruction, Lobo Well
were directionally drilled with casing using 83 was drilled with an S-shape trajectory. This borehole was drilled vertically to the kickoff point
steerable motors in a retrievable BHA. at 4,434 ft before building inclination angle to about 15 and then dropping back to a near-
vertical inclination after achieving about 500 ft of lateral displacement.
Lobo Well 83 included an interval that was
directionally drilled with 7-in. casing because of
a surface obstruction. The planned S-shape
trajectory called for building inclination to about
15 and then dropping angle back to near 13. Warren T and Lesso B: Casing Directional Drilling, at the SPE/IADC Drilling Conference and Exhibition,
paper AADE-05-NTCE-48, presented at the American Amsterdam, February 2325, 2005.
vertical after achieving sufficient lateral Association of Drilling Engineers (AADE) National Plcido JCR, Medeiros F, Lucena H, Medeiros JCM,
displacement to reach the subsurface target Technical Conference and Exhibition, Houston, Costa VASR, Silva PRC, Gravina CC, Alves R and
April 57, 2005.
(above right).16 Warren T: Casing DrillingExperience in Brazil,
Warren T and Lesso B: Casing Drilling Directional paper OTC 17141, presented at the Offshore Technology
This well was drilled vertically to the kickoff Wells, paper OTC 17453, presented at the Offshore Conference, Houston, May 25, 2005.
point at 4,434 ft [1,351 m], where the straight Technology Conference, Houston, May 25, 2005. 16. Strickler R, Mushovic T, Warren T and Lesso B: Casing
14. Warren et al, 2000, reference 9. Directional Drilling Using a Rotary Steerable System,
drilling assembly was retrieved by wireline and paper SPE/IADC 92195, presented at the SPE/IADC
15. Maidla E, Haci M, Jones S, Cluchey M, Alexander M and
replaced with a directional BHA that included a Warren T: Field Proof of the New Sliding Technology for Drilling Conference and Exhibition, Amsterdam,
Directional Drilling, paper SPE/IADC 92558, presented February 2325, 2005.
434-in. PDM. Drilling operations required
intermittent slide drilling from the kickoff point
to 4,808 ft [1,465 m] to build angle and establish
the desired direction.

Summer 2005 53
 The 434-in. steerable motor ran for only 154 ft
[47 m] before being replaced with a 512-in. motor
7-in. casing
to surface that generated higher torque at lower pressures
Vibration-
monitoring device and speeds (left).
When borehole angle reached about 10, the
Drill Lock well was drilled in rotating mode, which
Assembly (DLA)
increased the inclination angle to 15. The
borehole inclination could be increased easily,
but dropping angle required continuous sliding.
Slide drilling was reinitiated at 5,634 ft [1,717 m]
to return the trajectory closer to vertical. Even
MWD system
Tandem internal after switching to the larger PDM, a signicant
casing stabilizers number of stalls occurred, which required that
the motor be run at lower speeds and torque
loads during sliding (bottom left).
Casing shoe Drill collar spacer joint,
or float sub
Recovering from stalls while drilling with
casing was quicker than with drillpipe. The casing
was sufciently stiff, so reorientation was not
required. The bit was simply picked up to restart
the motor and then worked back to bottom to
continue drilling. Slide drilling without full string
rotation signicantly reduced the ROP, conrming
Nonmagnetic
drill collar PDM limitations reported in other wells.17
4 3/4-in. or 5 1/2-in. steerable
motor with 1.5 After the borehole inclination reached 10
bent housing again, the steerable motor assembly was pulled
and replaced with a rotary BHA. This pendulum
BHA was configured with the underreamer
immediately outside the casing and the
6 1/4-in. to 8 7/8-in.
underreamer directional control portion in the pilot hole.
Drilling with this assembly decreased the
6 1/4-in. PDC pilot bit
borehole angle from 10 to less than 2 of
inclination, which was maintained until the
> Lobo Well 83 retrievable BHA for a steerable downhole motor. The BHA assembly was pulled at 7,861 ft [2,396 m]
for drilling a directional interval with 7-in. casing in Lobo Well 83 included (next page).
tandem stabilizers inside the casing to reduce vibrations and wear on the The ROP was substantially higher while
DLA, a nonmagnetic drill collar, a vibration monitor, an MWD system, a
rotary drilling, even when limiting the WOB to
spacer, or oat, joint and a 434-in. motor with a 1.5 bent housing. The
assembly ended in an underreamer that opens up to 878 in. and a 614-in. ensure that borehole inclination decreased as
polycrystalline diamond compact (PDC) pilot bit. desired. A downhole vibration-monitoring device
recorded high lateral vibration while drilling
with this assembly, but relatively few motor stalls
occurred while drilling in rotary mode, and the
ROP improved signicantly.
2,400 The directional performance of this rotary
Motor stalls
2,200 assembly conrmed that borehole inclination
Pump pressure, psi

could be controlled in a small pilot hole even

2,000
with the underrreamer at a considerable
1,800
distance above the active portion of the BHA.
1,600 This test established confidence that RSS
1,400 technology could be used to drill with casing.
Unloaded motor pressure Currently, however, there are no RSS tools that
1,200
0 1 2 3 4 5 6 can work above an underreamer.
Time, hr
> Lobo Well 83 downhole motor performance. Slide drilling without full string rotation resulted in
frequent stalls during casing directional drilling with a steerable motor in Lobo Well 83.

54 Oileld Review
 y
Directional operations with casing and a
7-in. casing steerable PDM, especially in smaller hole sizes,
to surface are not efcient. It is easier to build inclination
6,000
than to drop angle with a smaller motor and
6,200 BHA. Even with drillpipe, orienting a PDM for a
Drill Lock directional correction can take several hours at
t
Assembly (DLA) 0 0f depths of 25,000 ft [7,620 m] or more. In addition
6,400 / 1
1.7 to numerous stalls, the ROP generally decreases

Measured depth (MD), ft

6,600 when using motors.
Using a steerable PDM demonstrated that it
6,800 is possible to drill directional wells with casing,
Tandem internal but drilling efciency during these trials was not
casing stabilizers 7,000
competitive with newer rotary steerable
technology, which now is used in about 60% of
7,200
directional wells drilled offshore.
7,400
Rotary Steerable Systems
7,600
Success in reducing lost circulation during the
Lobo drilling program sparked interest in
6 1/4-in. to 0 2 4 6 8 10 12 applying casing while drilling offshore, where
8 7/8-in. Wellbore inclination, degrees
underreamer
directional wells are a necessity. However, the
limitations of drilling directionally with casing
Stabilizer and steerable motors posed a problem. Rotary
steerable technology, developed to drill
directional, high-angle, horizontal and extended-
reach wells, appeared to be a viable alternative.
Nonmagnetic In many situations, rotary drilling with a RSS
drill collar is more efcient than using a downhole motor,
even for vertical, straight-hole applications.
Directional drilling with RSS technology
Stabilizer
eliminates orientation without rotation, or slide
drilling, making it possible to drill record-
Active
directional breaking distances, such as the extended-reach
control wells in Wytch Farm eld, UK, that are difcult
to drill with downhole motors.18
As RSS systems became more durable and
more reliable, they were deployed under
MWD system increasingly demanding conditions offshore.
Initially, RSS tools were applied primarily in
deepwater wells. However, as RSS efficiency
improved and their performance became better
known, costs decreased and companies
6 1/4-in. PDC pilot bit converted from steerable motors to RSS
technology for directional operations, especially
> Lobo Well 83 pendulum assembly and rotary drilling performance. After in the North Sea and the Gulf of Mexico.
reducing the inclination angle of Well 83 from 15 back to 10, the steerable
motor assembly was replaced by a pendulum assembly (left). This second BHA 17. Warren T, Tessari R and Houtchens B: Directional
with two stabilizers between the pilot bit and the underreamer, which was Casing while Drilling, paper WOCD-0430-01, presented
at the World Oil Casing Drilling Technical Conference,
positioned immediately below the casing, completed the drop in borehole Houston, March 3031, 2004.
angle back to near vertical. With the directional-control, or active, portion of 18. Meader T, Allen F and Riley G: To the Limit and
the BHA in the pilot hole, the driller was able to decrease the inclination angle BeyondThe Secret of World-Class Extended-Reach
from 10 to less than 2 (right). This directional performance conrmed that Drilling Performance at Wytch Farm, paper IADC/
well inclination could be controlled in the pilot hole while the underreamer SPE 59204, presented at the IADC/SPE Drilling Conference
enlarged the main hole at a considerable distance above the active part of the and Exhibition, New Orleans, February 2325, 2000.
BHA. In addition, the ROP increased signicantly while drilling in rotary mode
with this assembly.

Summer 2005 55
 Control unit

Plan View of Actuators

Bias unit

Actuators

Actuator
Applied
Directional force
correction

PDC bit

Drilling
tendency
> Rotary steerable technology. A rotary steerable system (RSS) applies force against the borehole
wall during full rotation of the entire drillstring to achieve a desired borehole trajectory. The
PowerDrive Xtra system, for example, comprises a control unit that houses electronics and
sensors (right). Based on commands from the control unit, the bias unit sequentially actuates
three external pads, which apply force against the borehole wall at the appropriate point during
each rotation to direct the bit in the required direction (bottom left). In vertical mode, this RSS tool
senses deviation away from vertical and automatically thrusts the bit back to vertical. Several
PowerDrive systems are available for drilling 412- to 1814-in. holes.

A rotary steerable system is ideal for
directional control in the retrievable BHA used
for drilling operations with casing. It minimizes
or eliminates many of the problems associated
with slide drilling, PDM performance limitations
and directional control difficulties, while
providing a smooth borehole that reduces torque.
Compact and mechanically uncomplicated RSS
tools are available for use in casing while
drilling (above).19
PowerDrive systems incorporate a bias unit
and a control unit in a 12.5-ft [3.8-m] housing.
The bias unit, located directly above the bit,
applies force in a controlled direction while the
entire drillstring is rotated from the surface. The
control unit, positioned behind the bias unit,
contains self-powered electronics, sensors and a
mechanism that applies a lateral force in the
specied direction required to achieve a desired
trajectory. The bias unit has three external,
hinged pads activated by controlled mud ow.
A three-way rotary disk valve sequentially
diverts mud into the piston chamber of each pad
as it rotates into proper alignment to apply force
in the direction opposite a desired trajectory. The
bit is constantly moved in one direction. If a
change in direction is not required, the system
operates in neutral mode, with each pad
extending sequentially to effectively push in all
directions and cancel each other.
During 2004, the Upstream Technology and
Lower 48 Exploration and Production groups at
ConocoPhillips began evaluating the feasibility of
using RSS tools in the pilot hole below the
underreamer for drilling operations with casing.20
This project represented the rst use of RSS
technology for casing directional drilling. The
challenge, however, was that there was little
overlap in logistics and methodologies for merging
casing while drilling with RSS technology.

56 Oileld Review
 Total length 94 ft [29 m]
Extension length 67 ft [20 m]
Weight in 5,200 lbm
drilling fluid [2,359 kg]

Tandem external
7-in. casing pilot-hole stabilizers
to surface

Drill Lock
Assembly (DLA)

8 joints of
7 5/8-in. casing

Tandem internal
casing stabilizers Drill collar

Casing shoe

Fluid filter

Drill collar spacer

joint, or float sub
> Lobo Well 89 retrievable BHA for vertical incli-
nation control. Vertical drilling operations with
7-in. casing required a RSS assembly with tandem 4 3/4-in. PowerDrive Xtra
6 1/8-in. to 8 7/8-in.
stabilizers inside the casing to dampen drilling underreamer 475 RSS
vibrations and to reduce wear and tear on the
DLA. A drill collar, or spacer sub, positioned the
underreamer outside the casing. External 6116-in.
stabilizers below the underreamer reduced
drilling vibrations in the pilot hole. A PowerDrive
Xtra RSS with a PDC bit completed the BHA.
6 1/8-in. PDC pilot bit

ConocoPhillips, Tesco and Schlumberger
conducted a two-well RSS test in the south Texas
Lobo trend using PowerDrive technology. The
rst RSS test with casing was conducted in a
vertical well. The second well was drilled
directionally with casing and a RSS.
Rotary Steerable Vertical Test
In June 2004, ConocoPhillips, Schlumberger and
Tesco performed the vertical RSS drilling test with
casing in Well 89, located about 30 miles [48 km]
northeast of Laredo, Texas. The vertical section
for surface casing was drilled to 588 ft [179 m]
using 958-in. casing and a retrievable BHA with an
812-in. pilot bit and a 1214-in. underreamer.
A review of 7-in. casing designs for drilling
vertical wells found that running heavyweight
758-in. integral flush-joint casing without
centralizers as the bottom eight joints reduced
drilling vibrations and fatigue failures. In
addition, engineers found that connections with
a beveled lower edge also reduced casing
vibration and wear.
After the surface casing was cemented in
place, a 434-in. PowerDrive Xtra 475 RSS
programmed to maintain a vertical borehole and
a 434-in. drill collar were added to the standard
BHA for 7-in. casing (above). This retrievable
19. Kuyken C: Rotary Steerable Technology: Pushing the
Limit, Oileld Review 16, no. 4 (Winter 2004): 1.
Copercini P, Soliman F, Gamal ME, Longstreet W, Rodd J,
Sarssam M, McCourt I, Persad B and Williams M:
Powering Up to Drill Down, Oileld Review 16,
no. 4 (Winter 2004): 49.
Brusco G, Lewis P and Williams M: Drilling Straight
Down, Oileld Review 16, no. 3 (Autumn 2004): 1417.
Williams M: Better Turns for Rotary Steerable Drilling,
Oileld Review 16, no. 1 (Spring 2004): 49.
Downton G, Hendricks A, Klausen TS and Patis D:
New Directions in Rotary Steerable Drilling, Oileld
Review 12, no. 1 (Spring 2000): 1829.
20. Strickler et al, reference 16.

Summer 2005 57
 BHA drilled to 4,821 ft [1,469 m] in 105 hours. ConocoPhillips switch back to drilling with The initial well plan called for building
Single-shot surveys taken every 500 ft indicated a drillpipe and a conventional BHA at considerable inclination angle to 29 and then dropping
near-vertical borehole inclination. Drilling additional expense. This dictated careful design, vertically into the target.
proceeded without problems, but engineers planning and implementation of the second test Unfortunately, wellhead and surface facilities
attributed higher than expected vibrations to the to fully evaluate casing directional drilling with for Well 79 were located between the remaining
long BHA extension. a RSS. open space for a rig and the subsurface target of
This run terminated at a planned Well 91. A new trajectory was designed to avoid
underreamer replacement. Drilling operations Rotary Steerable Directional Test colliding with the existing wellbore. This prole
continued to the 7-in. casing point at 7,620 ft Most wells in the Lobo development area are resembled well trajectories common on
[2,323 m]. ConocoPhillips retrieved the BHA, vertical. In late 2004, however, Well 91 presented multiwell offshore platforms (below right).
which was inspected and found to be in good a unique opportunity. The proposed location was Another factor complicated drilling
condition, and extracted operational data from about 1,200 ft [366 m] south of Well 79, a vertical operations. Well specications called for surface
the RSS tool. A multishot gyroscope run well that had been drilled with casing in March casing to be set at 1,270 ft [387 m]. The 958-in.
confirmed that the PowerDrive tool could 2004. The ConocoPhillips teams proposed using casing point for Lobo wells varies between 550
maintain verticality (below left). the existing surface location of Well 79 to and 2,400 ft [168 and 732 m], but experience
The vertical test conrmed RSS functionality directionally drill an S-shape trajectory with indicates that wells with deeper surface casing
and directional performance in a retrievable casing to reach the subsurface target for Well 91. have more problems with casing vibration and bit
assembly, and led to approval for a second test. This plan avoided building another location, instability, or whirl, during drilling of the 7-in.
In the next well, a more advanced BHA with an but the expense of directional operations was casing section because of casing-on-casing
MWD system and full directional capabilities more than three times the cost of a new rig pad. friction inside longer surface sections.
would be used to follow a planned trajectory. ConocoPhillips planned no other directional Adding a straight PDM above the
Inability to drill directionally or encountering wells for 2004, so this was the best option for underreamer addressed this problem, but
significant problems would require that testing casing directional drilling with a RSS. represented a significant change from the

0
Vertical Plot Horizontal Plot
0 200
1,000 RSS bias unit Surface
did not stabilize location
Total depth (TD)
Well 79 0
1,000
RSS vertical test
2,000
Well 79
-200
2,000
True vertical depth (TVD), ft

North-South departure, ft

3,000
True vertical depth (TVD), ft

Surface
location
-400
3,000
4,000

-600
4,000
5,000
RSS bias unit Well 91
fully operational -800
5,000 Well 91
6,000

-1,000
TD
7,000 6,000
Target
7-in. casing -1,200
section TD -600 -400 -200 0 200
8,000 7,000
0 0.5 1.0 1.5 2.0 2.5 3.0 0 1,000 2,000 East-West departure, ft
Wellbore inclination, degrees Horizontal displacement, ft
> Gyroscope survey for the 7-in. casing section in > Lobo Well 91 vertical and horizontal trajectory plots. To avoid risk of a
Lobo Well 89. Inclination data from Lobo Well 89 collision with Well 79, the horizontal trajectory of Well 91 started off along
indicated that the 434-in. PowerDrive Xtra 475 an azimuth 40 east of the target azimuth before beginning a 100 right turn
control unit did not stabilize on a gravity to the southwest (right). The vertical trajectory built inclination angle to
reference until the bit reached 3,710 ft [1,131 m]. 29 (left). During later stages of the horizontal turn, drillers initiated a drop
The well drifted to an inclination angle of 2.25 at in angle to bring the borehole into the target at a near-vertical inclination.
3,600 ft [1,097 m]. The RSS regained full function This prole resembled those used on large offshore platforms with multiple
and directional control from 3,710 to 4,821 ft. At drilling slots.
3,800 ft [1,158 m], the borehole trajectory returned
to a near-vertical inclination of 0.25 for the
remainder of the RSS vertical test. There was a
slight tendency to build angle from about 2,000 ft
[607 m] to 3,800 ft when the RSS tool was not
effective and again after the test ended at 4,821 ft.

58 Oileld Review
 < Lobo Well 91 retrievable BHA for rotary direc-
Total length 112 ft [34 m]
tional drilling. Directional operations with 7-in.
Extension length 85 ft [26 m] casing and a RSS required several innovative
Weight in 6,200 lbm BHA components. Tandem stabilizers inside the
drilling fluid [2,812 kg] casing dampened drilling vibrations and helped
protect the DLA. A straight 6-in. PDM served as a
spacer sub and added rotation to the BHA and bit
so that surface rotation of the drillstring could be
Tandem external decreased when dealing with high drilling vibra-
7-in. casing pilot-hole stabilizers tions. A jet nozzle below the underreamer
to surface diverted 20% of the drilling uid from the bit to
balance ow between the 618-in. pilot hole and
the 878-in. main borehole. External 6116-in. stabiliz-
ers in the pilot hole below the jet nozzle reduced
Drill Lock vibration and wear on the underreamer. A slim
Assembly (DLA) MWD system and a PowerDrive Xtra RSS with a
PDC bit completed the BHA.

8 joints of
7 5/8-in. casing

MWD system
Tandem internal
casing stabilizers

Casing shoe
Fluid filter

6-in. straight
PDM

4 3/4-in. PowerDrive Xtra

475 RSS
6 1/8-in. to 8 7/8-in.
underreamer

Jet nozzle

6 1/8-in. PDC pilot bit

vertical test in Well 89. The purpose of this motor
was to allow reduced surface rotation of the
drillstring when dealing with excessive
vibrations. The motor also protected the drilling
string and BHA by acting as a shock absorber.
However, the MWD system had to be run below
the motor, so the MWD signal had to travel up
through the motor. This was technically feasible,
but had never been done (above).
Casing directional drilling requires bit
rotational speeds that are similar to drilling with
drillpipe, typically 120 to 180 rpm. The motor
added rotation back into the BHA and bit to
maintain an adequate ROP. For example,
if bit whirl limits surface rotation to 50 rpm, the
motor adds 100 rpm to reestablish optimal
bit performance.
The underreamer, which opened the 618-in.
pilot hole to 878 in., was placed directly below the
mud motor. A jet nozzle diverted 20% of the
drilling fluid from the bit to balance flow
between the pilot hole and the expanded
borehole. Tandem 6116-in. external stabilizers
were positioned below the jet nozzle to reduce
vibration and wear on the underreamer. A
PowerDrive Xtra 475 RSS and a 618-in. PDC bit
were installed below the MWD system.21
21. Downton GC and Carrington D: Rotary Steerable Drilling
System for the 6-in Hole, paper SPE/IADC 79922,
presented at the SPE/IADC Drilling Conference,
Amsterdam, February 1921, 2003.

Summer 2005 59
 Drilling operations with 7-in. casing began at
1,278 ft [390 m]. A four-blade PDC bit with 34-in.
cutters was used to drill this section, the same
type of bit used in other Lobo wells. Surveys
indicated that the borehole was nearly vertical.
The MWD system located below mud motors
maintained reliable data transmission. However,
surveys had to be taken during quiet periods
when rig pumps were off and there was no motor
rotation, instead of when the pumps rst came
back on after a casing connection, which is
common practice. Signal attenuation of MWD
telemetry through the motor was only 40 to 50%
instead of the expected 90%.
After the kickoff depth of 2,100 ft [640 m] was
reached, the build section was completed as
planned. The initial run continued to 4,067 ft
[1,240 m], where pressure spikes indicated a
problem, so the BHA was retrieved by wireline.
The motor had locked up and there was a
washout, or hole, in the RSS tool, but it was still
operational. A PDM was not rerun. The bias unit
of the RSS was replaced and drilling continued.
Drilling was slower, and it was difcult to keep
surface rotation above 60 rpm without the motor.
This second run ended when a replacement
motor arrived on location. The motor was added
back for the third run, restoring the BHA to the
initial design conguration. Drilling proceeded for
200 ft [61 m] before the ROP dropped signicantly.
When the BHA was pulled, drillers found that the
small stabilizer under the underrreamer cutter
pads was larger than the bit, 614 in. instead of 618 in.
This oversized stabilizer worked until harder
formations were encountered.
The underreamer was replaced, and drilling
continued without incident until reaching 5,420 ft
[1,652 m], where the casing became differentially
stuck. Directionally, the build and turn sections
were completed, and the drop back to vertical
was under way. Drilling continued to 6,360 ft
[1,939 m]. The two instances of nonproductive
time in directional Well 91, an oversize stabilizer
and stuck pipe, added about 85 hours to the total
drilling time.
The borehole was now at a 4 inclination
angle. A pressure drop indicated a washout in the
BHA. Surface inspection revealed a washout in
the connection between the jet nozzle and the
external tandem stabilizer. The jet nozzle was
removed from the BHA and drilling continued to
TD at 6,950 ft [2,118 m].
60
Using casing for drilling improves operational
efciency by eliminating pipe trips and reducing
unexpected difficulties encountered when
running casing in a separate operation.
ConocoPhillips experience in Well 91 proved that
RSS technology is effective for casing directional
drilling in smaller 812 to 978-in. hole sizes where
PDM performance is limited (next page).
Bit selection issues common in directional
drilling with conventional drillpipe and a RSS
must be addressed to drill directionally with
casing. Bits are chosen based on their side-
cutting capability for directional control and
their stability to reduce excessive vibrations. Bit
hydraulics and BHA nozzles also have to be
balanced so that uid ow rates in both the pilot
hole and the full-gauge borehole remain within
optimal ranges for effective bit and hole
cleaning, and for operating MWD systems and
PDM or RSS tools.
If the borehole surface is irregular or rough
and the well path is tortuous, casing stiffness can
contribute to higher torque. Lateral and
torsional forces are higher than with drillpipe
because larger tubulars weigh more and have a
greater rotating diameter. Casing string designs
for drilling directional wells require more
centralization than in vertical wells.
In addition, casing centralization plays an
important role in effective hole cleaning, and in
reducing drillstring vibrations and incidents of
pipe sticking. Hole cleaning and differential
sticking increase in directional wells with higher
inclination angles. Care must be taken to avoid
long periods of time when either the casing or
the BHA is stationary without uid circulation.
Casing while drilling, and to a greater extent
casing directional drilling, are still in early stages
of development. Procedures and practices will be
optimized as operator experience with these new
technologies increases.
An Expanding Range of Applications
Operators in the USA and Canada have drilled
commercial vertical wells with casing sizes ranging
from 412 in. to 1338 in. The deepest well drilled to
date was just over 13,000 ft [3,959 m]. Directional
wells have been drilled with casing and steerable
motors, but success is difcult to achieve in hole
sizes of less than 812 in. because a smaller PDM
supplies suboptimal torque for drilling.
Experience gained from vertical and
directional testing of rotary steerable technology
while drilling with casing proved that a 434-in.
RSS can effectively drill 812-in. holes with 7-in.
casing. Directional control in the pilot hole is
sufcient to guide larger diameter underreamers
and casing to a directional target. Schlumberger
is currently conducting eld trials of a 314-in.
ultraslim RSS for drilling with 6-in., 512-in. or
5-in. casing.
Acquiring well logs for formation evaluation
is a key consideration when evaluating casing
while drilling. Because casing remains in the
wellbore after reaching TD, operators must
identify the best methods for logging these wells
to take full advantage of casing while drilling and
its capabilities to reduce nonproductive rig time.
Currently, there are four options: run
conventional openhole wireline logs, run memory
logging tools in a retrievable BHA, run an LWD
system in the drilling BHA, or run new wireline
logging systems that acquire measurements
behind pipe.
To run openhole or memory logs, the casing
must be pulled into the previously cemented
casing. The casing has to be pulled above the
zones of interest, but does not have to be tripped
completely out of the well. If a kick occurs during
logging, it can be circulated out down to the top
of the openhole section. However, if the borehole
collapses, it may not be possible to acquire a log
across the entire interval.
Memory logs are acquired as the casing is
pulled back into the preceding casing string by
deploying logging tools in a retrievable BHA after
retrieving the drilling assembly. This approach
ensures that the entire openhole section can be
logged and evaluated. Continuous fluid
circulation keeps logging tools cool and reduces
the chance of a kick during logging.
LWD tools have been used in vertical wells
during drilling operations with casing,
eliminating the need to pull casing before
logging. However, the addition of LWD tools to a
retrievable BHA adds cost, weight and length,
which must be balanced against wireline
retrieval risks and vibration problems in longer
BHA extensions.
New technology now makes logging behind
casing possible. Schlumberger ABC Analysis
Behind Casing services are a cost-effective
alternative to openhole, memory or LWD
formation evaluation, allowing operators to
minimize nonproductive rig time by assessing
potentially productive intervals after reaching
TD without pulling or manipulating the casing. In
addition to acquiring resistivity, porosity, sonic,
bulk density, lithology, pulsed neutron and
reservoir pressure measurements behind
cemented pipe, ABC services also include
sampling of formation uids.22
Oileld Review
 0

9 5/8-in. surface casing TD

1,000

2,000 Rotary drill

to kickoff point

3,000
Measured depth (MD), ft

Kickoff point

4,000 Replace 4 3/4-in. PDM with 5 1/2-in. PDM

5,000 Drop angle Replace steerable

PDM with
rotary BHA
Well 79
Well 91
6,000

Well 83
7,000
7-in. production casing TD

8,000
0 2 4 6 8 10 12 14 16 18 20
Time, days
> Drilling time versus depth for Lobo Well 91, Well 79 and Well 83. Directional Well 91 (blue) and nearby vertical
Well 79 (red) were comparable over about 4,500 ft [1,372 m]. A total of 132 casing joints were used to drill
directionally in Well 91 compared with 128 joints for vertical Well 79. The ROP on a joint-by-joint basis for the
directional well was only about 10% less than the ROP in the vertical well. The trajectory was more complex in
Well 91, but drilling with casing and a RSS saved a substantial amount of time compared with Well 83 (black),
which was drilled using casing and a steerable PDM.

The capability of drilling directional wells
makes casing while drilling attractive for
offshore applications in areas prone to lost
circulation and previously uneconomic to drill
using conventional processes and techniques.
Modications of current systems are under way
to allow casing while drilling in deepwater
applications. Most deepwater casing strings are
set as liners. Several strategies are under
development to apply retrievable BHA
experience to liner drilling.
There are several potential applications that
require additional advances in equipment and
techniques. Research and development are
under way to allow underbalanced drilling using
casing and drilling with air. An obvious advantage
of using casing for air and underbalanced drilling
is that wells do not have to be balanced with
heavier mud, or killed, to trip drillpipe out of
the hole.
In the future, this technique may be used to
drill high-pressure, high-temperature (HPHT)
and geothermal wells. The combination of casing
while drilling and expandable tubulars ultimately
may provide a unique well-construction solution,
but additional hurdles must be overcome for this
to be practical. As casing directional drilling
becomes more common, market pressures will
likely stimulate the development of additional
systems and technologies specically for use in
casing-while-drilling applications. MET

22. Aulia K, Poernomo B, Richmond WC, Wicaksono AH,
Bguin P, Benimeli D, Dubourg I, Rouault G, VanderWal P,
Boyd A, Farag S, Ferraris P, McDougall A, Rosa M
and Sharbak D: Resistivity Behind Casing,
Oileld Review 13, no. 1 (Spring 2001): 225.
Bellman K, Bittner S, Gupta A, Cameron D, Miller B,
Cervantes E, Fondyga A, Jaramillo D, Pacha V, Hunter T,
Salsman A, Kelder O, Orozco R and Spagrud T:
Evaluating and Monitoring Reservoirs Behind Casing,
Oileld Review 15, no. 2 (Summer 2003): 29.

Summer 2005 61

Azhar Ali is based in Kerteh, Malaysia, as PETRONAS
Carigali Sdn Bhd (PCSB) Well Services Engineer. His
responsibilities include implementing, monitoring and
supervising the execution of well data acquisition, well
integrity implementation and rectication programs,
and production enhancement activities. He is charged
with protecting wells from operational hazards and
enabling prolonged, continuous and safe production.
Prior to joining PCSB in 2001, Azhar worked with BJ
Services as assistant cementer in the ExxonMobil eld
at Kerteh. He is a graduate of the University of
Technology Malaysia in Johor Bahru, with a degree
(Hons) in petroleum engineering.
Dan Barson is Senior Geological Specialist for
OILEXCO Inc. in Calgary. His responsibilities include
hydrodynamic and capillary pressure analysis, pay
mapping and prospect evaluation. Before joining
OILEXCO in 2005, he spent two years as an indepen-
dent consultant for Earth Science Consulting Inc.,
working on exploration geology and acquisition and
divestiture projects in the Gulf of Mexico and the
North Sea. Prior to that he spent 10 years with a
Calgary-based consultancy, developing a hydrodynamic
oil-migration modeling program, and applying it to
projects in Canada, South America, North Africa and
the North Sea. Dan holds a BS degree from Kings
College, University of London, and a PhD degree from
the University of Alberta, Edmonton, Canada, both
in geology.
Craig H. Bivins has served as President of Bivins
Energy Corporation, Dallas, since 1980, and is cur-
rently engaged in the assembly, management and
operation of exploratory and development programs
and prospects located primarily in the East Texas
basin. These projects include a large Jurassic, Cotton
Valley Sand gas development program in Harrison
County, a Lower Cretaceous, Rodessa/Pettit oil and gas
exploratory program located on the Chandler Ridge
Turtle structure in Henderson and Smith Counties and
a Jurassic, Bossier Sand gas development program in
Freestone and Navarro Counties. In addition, Craig
has also participated in projects in West Texas, New
Mexico, Oklahoma, Louisiana and the Texas Gulf
Coast. He holds a BBA degree in nance, marketing
and real estate from Southern Methodist University
in Dallas.
Curtis G. Blount is a Well Supervisor and Adviser
specializing in remedial well interventions, and is
also a Coiled Tubing (CT) Technology Adviser for
ConocoPhillips Alaska Inc. in Anchorage. He has been
active in CT research and applied technology develop-
ment for more than 20 years. Curtis has coauthored
more than 30 technical papers and holds 20 patents.
He served as an SPE distinguished lecturer on CT
drilling technology, cochaired the rst SPE forum
on CT technology, twice cochaired the SPE/ICoTA
Roundtable and presented the SPE Applied
Technology Workshop on Coiled Tubing Drilling.
He serves on the Journal of Petroleum Technology
editorial committee.
62
Contributors
Curtis Boney is Business Development Manager for
Schlumberger Well Production Services and Coiled
Tubing Services in the North and South America
(NSA) GeoMarket* region. Based in Sugar Land,
Texas, USA, he supports coiled tubing and stimulation
solutions for NSA clients. Curtis joined the company,
then Dowell, in 1974 as a service sales engineer,
supervising stimulation and cementing jobs. Since
then, he has worked in many cementing and stimulation
technical support and management positions through-
out Texas and New Mexico, USA. He earned a BS
degree in agricultural engineering from Texas Tech
University in Lubbock.
Rod Christensen, Vice President, Exploration, for
OILEXCO Inc., in Calgary, is currently developing a
drilling portfolio concentrated in the North Sea, with
several exploration drilling prospects scheduled in
2005. He is also working on drilling plans for three
horizontal wells in the Brenda pool. Since 1993, he
served as president of Cuesta Energy Inc., where he
provided consulting and geological services to various
petroleum companies before joining OILEXCO in 2005.
Rod has been credited with discoveries including the
Kisbey eld, the Heward eld, the North Browning
eld, the Carlyle North pool and the Steppe/North
Manor pool during his 26-year career. He has a BA
degree in zoology and a BS degree in geological sciences
from the University of Washington in Seattle, USA.
Eric Decoster obtained an engineering degree at
Ecole Centrale in Paris and an MS degree in civil engi-
neering at the University of Wisconsin, Madison, USA.
He joined Schlumberger as a eld engineer in the
Middle East in 1978. In 1996, after assignments in
log interpretation and marketing in various locations
around the world, Eric became principal petrophysicist
for the Venezuela-Trinidad area, based in Caracas.
In this role, he oversees petrophysical support for all
Schlumberger wireline services, focusing primarily
on nuclear magnetic resonance, spectroscopy and
resistivity logging.
John Engels, who is based in Houston, is a Schlumberger
Senior Stimulation Technical Representative for
Anadarko Petroleum, The Woodlands, Texas. He works
with all Anadarko asset and production engineering
teams on stimulation concerns and sales. He joined
the company in 1996 as a eld engineer specializing in
fracturing operations in Hassi Messaoud, Algeria. He
remained there as district technical fracturing engi-
neer until transferring to Houston as product champion
for fracturing uids in 2000. Before assuming his
current position in 2004, he worked in Sugar Land,
Texas, as technical support engineering director and
senior stimulation support engineer. John holds a BS
degree in civil engineering from the University of
South Florida, Tampa, USA.
Eugene O. Fielder, who works for Devon Energy, is
based in Oklahoma City, Oklahoma, USA. He is
Operations Engineering Supervisor for the Fort Worth
basin in northern Texas. He joined Devon in 2002 as
reservoir engineering supervisor after many years in
the industry with Mitchell Energy and has worked in
reservoir and production engineering in the Barnett
Shale eld since 1995. Eugene received a BS degree in
petroleum engineering from Texas A&M University,
College Station.
Kyle R. Fontenot is Operations Manager for
ConocoPhillips in Venezuela. He earned BS and MS
degrees in petroleum engineering from Louisiana
State University and has 18 years of drilling experience
with BP and ConocoPhillips in Gulf of Mexico, New
Zealand, Venezuela, Indonesia, Norway, North Sea and
Texas operations. Kyle has worked as a supervisor and
site engineer on land, platform, jackup, posted barge,
submersible, semisubmersible, deepwater drillship
and casing-drilling rigs. Previously, he was lead engi-
neer for the Deepwater Technology Drilling Team and
well operations coordinator for the South Texas
Business Unit.
Chris Fredd, North and South America (NSA)
Stimulation Products Manager for Schlumberger
in Sugar Land, Texas, provides technical support for
eld operations and clients and manages the NSA
Stimulation Client Support Laboratory. He focuses on
solving client-related problems, addressing competitive
issues, evaluating and introducing new technology and
supporting service quality standards. He is also respon-
sible for providing technical training to laboratory and
eld personnel. Chris joined Schlumberger as a staff
engineer in 1997, working on improved cleanup of frac-
turing uids and evaluation of acidizing technology. He
later worked as an area laboratory manager for district
laboratories in south Texas. He obtained a BS degree
from Clarkson University, Potsdam, New York, USA, and
MS and PhD degrees from University of Michigan in
Ann Arbor, USA, all in chemical engineering.
Tim Gorham is a Petroleum Engineer for Chevron
North America Exploration and Production based in
McKittrick, California, USA. He is currently a production
engineer for the Cymric Technical Team. Tim has
worked in the San Joaquin Valley primarily in produc-
tion, completion and stimulation of light oil and gas.
He holds BS and MS degrees in petroleum engineering
from New Mexico Tech in Socorro.
Jim Grau works at Schlumberger-Doll Research,
Ridgeeld, Connecticut, USA, on the various problems
of extracting information from gamma ray spectrometry
logging tools. He joined Schlumberger in 1977 after
receiving a BS degree in engineering physics from the
University of Toledo, Ohio, USA, and MS and PhD
degrees in low-energy nuclear physics from Purdue
University, West Lafayette, Indiana, USA.
Udit Kumar Guru works as the Schlumberger
Petrophysics Domain Champion for the East Africa
and East Mediterranean GeoMarket region. Based in
Cairo, he supports petrophysical interpretation needs
and introduces new technologies for clients. Before
moving to Cairo, he was domain champion for the
India GeoMarket region, supporting interpretation for
deepwater operations. Udit earned an MS degree in
exploration geophysics from the Indian Institute of
Technology, Kharagpur. He then spent 15 years as a
petrophysicist with an exploration and production
company before joining Schlumberger.
Oileld Review
Michael Herron is a Scientic Advisor working on
applications of geochemical and statistical methods
for reservoir interpretation problems at Schlumberger-
Doll Research in Connecticut. Before joining
Schlumberger in 1982, he studied the chemical
stratigraphy of polar ice cores as part of his doctoral
work at the State University of New York at Buffalo,
where he received a PhD degree in geological sciences.
Mike also has a BA degree in chemistry from the
University of California, San Diego.
Susan Herron has served as Program Manager, Nuclear
Program, in the Sensor Physics department at
Schlumberger-Doll Research in Connecticut, since
1998. The program focuses on the development of new
technology and measurement systems, and nuclear
modeling with emphasis on nuclear spectroscopy and
sourceless logging. Since joining the company in 1984,
she has worked primarily on development of applica-
tions for nuclear spectroscopy. She developed interpre-
tation techniques for quantifying lithology and matrix
properties from elemental concentrations and for inte-
grating nuclear spectroscopy with conventional logs for
rapid formation evaluation. Susan holds a BA degree in
geology from Tufts University, Medford, Massachusetts,
USA, and MS and PhD degrees in geological sciences
from the State University of New York at Buffalo.
Stephen Hill, based in Sugar Land, Texas, is a
Schlumberger Senior Engineer working for Coiled
Tubing Services in the CoilSOLUTIONS group where he
is currently responsible for tool development on the
BridgeFRAC project. He continues to support
PowerCLEAN* engineered ll removal service and is
the coiled tubing representative for both the Tulsa
University Drilling and Research Projects and the
Coiled Tubing Consortium at the University of
Oklahoma. Before joining the company in 2000,
Stephen worked as a graduate research assistant at
Georgia Institute of Technology, Atlanta, USA, and as
an engineer for the National Institute of Standards
and Technology, Gaithersburg, Maryland, USA. He
received a BS degree in general science from
Morehouse College in Atlanta, and BS, MS and PhD
degrees in mechanical engineering from the Georgia
Institute of Technology.
Martn Jordn is a Petrophysicist in the Reservoir
Department of the Southern Central Division of
Petrleos de Venezuela S.A. (PDVSA), where his
responsibilities include technical support in petro-
physical and geological evaluation of wells drilled in
the Apure basin, including the Guata eld. He gradu-
ated as an engineer in geology from the Universidad
de Oriente in Ciudad Bolivar, Venezuela, and began
his career as a wireline eld engineer and log analyst
for Halliburton in Las Morochas, Venezuela. In 2001,
Martn joined the Reservoir Department of PDVSA in
Barinas, Venezuela, where he has promoted the use
of numerous new technologies, in both open- and
cased-hole wells, to improve reserves evaluation and
ultimate recovery.
Summer 2005
Tobias Judd, based in Mexico City, is Schlumberger
Well Services Technical Engineer for the Mexico and
Central America (MCA) GeoMarket area. He is respon-
sible for integrated hydraulic fracturing initiatives and
production optimization with PEMEX (Petrleos
Mexicanos) operations. He also provides technical
support for well services contracts. Tobias joined the
company in 1997 as a stimulation eld engineer in El
Tigre, Venezuela. He also worked in various locations
in Argentina and Brazil as engineer-in-charge, DESC*
design and evaluation services for clients fracturing
engineer and eld service manager before taking his
current post in 2004. He is a graduate of the University
of Colorado, Boulder, USA, with a BS degree in chemi-
cal engineering.
John Lassek is Product Development Manager for
Schlumberger in Sugar Land, Texas. His responsibilities
include managing market introductions of new prod-
ucts and services, focusing on developing business
plans, and marketing, training and documentation for
each product. After receiving a BS degree in petroleum
engineering from the University of Alaska, Fairbanks,
USA, he joined the company in West Texas as a eld
engineer performing cementing and stimulation
operations. He worked in Indonesia and the Java Sea
as a stimulation specialist before transferring to
Oklahoma as a district technical engineer and then as
regional engineering manager responsible for cement-
ing, stimulation and coiled tubing projects for US Land
East. He also worked as technical projects leader for
Schlumberger Data and Consulting Services and was
responsible for evaluation of well completions.
In 2004, John transferred to Houston as a product
champion for innovative uid products before taking
his current post.
Bill Lesso, Casing Drilling Advisor, works for
Schlumberger Drilling and Measurements in Houston.
He is currently working with ConocoPhillips in deploy-
ing casing directional drilling in Norway and China.
He joined the company in 1976 as a eld engineer in
Dayton, Texas, after receiving a BS degree in mechani-
cal engineering from the University of Texas at Austin.
Since then, Bill has held positions in wireline, well ser-
vices and drilling and measurements. He has managed
horizontal and geosteering projects in Malaysia, UK,
Latin America and several USA locations. He became
involved in casing drilling after lling a visiting posi-
tion at Schlumberger Cambridge Research, England.
M. J. Loveland is a Well Integrity Supervisor for
ConocoPhillips in Kuparuk, Alaska. She has worked in
the industry for 15 years, eight of which have been in
Kuparuk, in various operations analytical engineering
and production engineering assignments. She earned a
BS degree in petroleum engineering from University of
Wyoming, Laramie, USA.
Thomas M. Maher, Geological Manager, Apache Egypt
Companies, is based in Cairo. He is responsible for
managing geological efforts in prospecting, operations
and petrophysical evaluation for Apache's active
exploration program in the western desert of Egypt. He
joined Apache Corporation in 1986 and, before taking
his current position in 2002, worked in Tulsa as a staff
geologist, geoscience manager and exploration and
development manager for the USA mid-continent
region. Thomas has a BS degree in geology from the
University of Massachusetts at Amherst, an MS degree
in geology from Miami University in Oxford, Ohio, and
an MBA degree from the University of Phoenix in Tulsa.
Shahril Mokhtar is District Technical Engineer for
Schlumberger, based in Abu Dhabi, United Arab
Emirates (UAE). Before taking this position in June
2005, he was the DESC engineer for PETRONAS
Carigali Sdn Bhd (PCSB) in Kemaman, Malaysia.
He was responsible for PCSB and Carigali Triton
Operating Company technical and commercial
projects including PowerCLEAN and LiteNET coiled
tubing systems. He joined Schlumberger in 2001 as a
coiled tubing eld engineer and worked in Malaysia,
Indonesia, UAE, Scotland and Thailand. Shahril
earned a BE degree (Hons) in mechanical and
aeronautical design engineering from the University
of Brighton, England.
Jessica Pedota is a General Field Engineer and DESC
engineer for Schlumberger Well Intervention Services,
Anchorage, where she writes procedures for well
integrity work and other well operations including
e-line, cementing and coiled tubing. She also provides
technical assistance for coiled tubing operations for
BP Prudhoe Bay, Alaska operations. Jessica joined the
company as a eld engineer in 2002 at Prudhoe Bay
after internships in Prestonsburg, Kentucky, USA,
and Prudhoe Bay/Kenai. She holds a BS degree in
chemical engineering from Michigan Technological
University, Houghton.
Jai Pokhriyal is the PowerCLEAN Product Champion
for Schlumberger in Sugar Land, Texas. He began his
career as an offshore and onshore drilling engineer
with Oil and Natural Gas Corporation Ltd., India, and
later worked as a research assistant at the University
of Wyoming in Laramie. He has a BS degree in
mechanical engineering from Maulana Azad College of
Technology, Bhopal, India, and an MS degree in petro-
leum engineering from the University of Wyoming.
Jai joined Schlumberger as a eld engineer in 1997
in Alice, Texas, where he designed, executed and
evaluated cementing services. He later worked as
cementing cell leader and as eld service manager,
offshore cementing, before moving to well production
services and coiled tubing in 2004.
Mads Rdsj, based in Stavanger, works as Drilling
Engineer, Well Integrity Delivery Team Leader for BP
Norge AS. He is responsible for delivering and design-
ing all casing and liner strings on the Valhall Flank
development (VFD) in the North Sea. Mads joined
the company in 2002 after earning an MS degree in
geosciences and petroleum technology from the
Norwegian University of Science and Technology
in Trondheim. Before taking his current position,
he worked as an offshore drilling engineer, an
onshore CT job ofcer on the VFD and an offshore
completion engineer.
63
Radovan Rolovic is Principal Engineer for
Schlumberger Stonehouse Technology Center in
Stonehouse, England, where he works on development
and manufacturing of rotary steerable advanced
drilling systems for downhole applications. He joined
the company in 1997 as a senior engineer and project
leader for the Dowell Coiled Tubing Product
Development group in Rosharon and Sugar Land,
Texas. He worked on development of new products,
including CoilCAT* coiled tubing computer-aided
treatment, CoilCADE* coiled tubing design and
evaluation software, PipeSAVER* coiled tubing storage
inhibition system and PowerCLEAN service. He trans-
ferred to Schlumberger Drilling and Measurements in
2004. Radovan holds a BS degree from the University
of Montenegro in Podgorica, an MS degree from the
University of Belgrade, Serbia & Montenegro, and a
PhD degree from the University of Tulsa, all in
mechanical engineering.
Erik Rylander, Schlumberger Product Champion for
the DecisionXpress* petrophysical evaluation system,
is based in Clamart, France. He joined the company in
1995 as a junior eld engineer in Duncan, Oklahoma,
and then moved to Equatorial Guinea and Nigeria as a
eld engineer (1996 to 1997). He spent the next four
years as an MDT* Modular Formation Dynamics
Tester specialist eld engineer with Gulf Coast
Special Services. Before taking his current position in
2004, he served three years as eld service manager
in that location. Erik obtained a BS degree in engi-
neering with an electrical engineering specialty from
Colorado School of Mines in Golden.
Alfredo E. Sanchez Mogollon is a Schlumberger DESC
engineer, focusing on fracturing services, for PEMEX
Exploracin y Produccin and Petrleo Brasileiro SA
(PETROBRAS) in Reynosa, Mexico. He provides in-
house support including design and evaluation of
fracture stimulation treatments, periodic service
quality reviews and tracking and database maintenance.
He began his career with Schlumberger Well Services
in 2000 as a eld engineer in Rock Springs, Wyoming.
He worked on uids rheology, laboratory analysis and
coordinated fracturing crews as Well Production
Services cell leader before transferring to Reynosa
in 2003. Alfredo earned a BS degree in materials
science and engineering from Universidad Simn
Bolvar in Caracas.
Chakib Sbiti, Executive Vice President of
Schlumberger Oileld Services (OFS), manages oil-
eld technology development and all operations in
the OFS business segment worldwide. Prior to assum-
ing this position in 2003, he was president, Middle
East and Asia, Schlumberger OFS for two years. He
also served as director of personnel, Oileld Services,
Paris; and vice president, Wireline & Testing, Europe,
Africa and the Mediterranean. Chakib joined
Schlumberger in 1981 as a eld engineer after
studying electrical engineering in France.
R. D. (Bob) Strickler, who is based in Houston, is
Staff Drilling Engineer for ConocoPhillips South Texas
Business Unit. Currently, he is involved in planning
and operations for ConocoPhillips Casing Drilling
Program. He joined Conoco in 1988 and has 27 years
of experience in planning, supervising and managing
production and drilling operations. He has worked
in offshore, shallow water and land areas in the
USA and internationally. Bob has a BS degree in
petroleum technology.
64
Phil Sullivan, Principal Engineer for Schlumberger
Well Services in Sugar Land, Texas, is working on
improved uid loss control for fracturing uids and
collaborating with researchers at Schlumberger
Cambridge Research, England, and at Princeton
University, New Jersey, USA. Since joining the com-
pany in 1994 as a development engineer for Dowell,
he has worked on uid systems, ber-assisted trans-
port uids, high-temperature ClearFRAC* polymer-
free frac uid and coiled tubing cleanout operations.
Phil has a BS degree from the University of Virginia,
Charlottesville, USA, and MS and PhD degrees from
Purdue University, West Lafayette, Indiana, all in
mechanical engineering.
Lloyd Tabor is Schlumberger Account Manager for
Devon Energy in Oklahoma City, Oklahoma. There he
is responsible for initiating, supporting and network-
ing all Schlumberger business in Devons western and
central US divisions. After earning a BS degree in
chemical engineering from Louisiana State University,
Baton Rouge, he joined Schlumberger, then Dowell, as
a eld engineer in Lafayette, Louisiana. He worked in
various locations as development engineer, operations
manager and oileld services senior sales engineer
before assuming his current post. During his career,
Lloyd has worked on cementing, fracturing and new
technology implementation and was responsible for
introducing CoilFRAC* stimulation through coiled
tubing service and ClearFRAC uids to Equitable
Resources and Columbia Energy in the eastern USA.
Ariel Valenzuela Muoz works for PEMEX in Reynosa,
Mexico, where he is in charge of the Fracturing
Optimization department. After receiving a degree in
petroleum engineering from Instituto Politcnico
Nacional in Mexico City, he joined the company in
1985 to work in petroleum engineering, production,
workover and well completions. Prior to taking his
current position, Ariel was assigned to the design and
perforation department where he was responsible for
the design, supervision and evaluation of hydraulic
fracturing treatments for the Burgos basin.
Tommy M. Warren is the Director, Casing Drilling
Research and Engineering, for Tesco Corporation in
Houston. He joined Tesco in 1999 after 26 years with
Amoco in operations and drilling research. His
research in roller-cone bit mechanics, drag bit
mechanics, directional drilling, drillstring mechanics,
high-speed drilling systems, rock mechanics and the
Casing Drilling system has led to the publication of
60 technical papers and 35 patents. Tommy earned
BS and MS degrees in mineral engineering from the
University of Alabama, Tuscaloosa, USA, and was
selected as a University of Alabama Distinguished
Engineering Fellow in 1994. He served as chairman
of the SPE 1999 Annual Technical Conference and
Exhibition, was an SPE Distinguished Lecturer in
1999 and is currently the chairman-elect of the SPE
Publications coordinating committee and member of
the IADC/SPE Drilling Conference program commit-
tee. He was the recipient of the 1997 SPE Drilling
Engineering Award.
Xiaowei Weng is Schlumberger Senior Development
Engineer in the Engineering Applications department,
in Sugar Land, Texas. He is responsible for engineering
and modeling support for software development in the
areas of hydraulic fracturing, acid fracturing and
coiled tubing cleanout. Prior to joining the company
in 1999, he worked as a senior engineer in fracturing
technology and eld design support for ARCO E&P
Technology. Xiaowei holds a BS degree from the
University of Science and Technology of China, Hefei,
and MS and PhD degrees from the University of Texas
at Austin, all in engineering mechanics.
Jim White is a Petrophysical Advisor for
Schlumberger UK, based in Aberdeen. He provides
interpretation support for new wireline sensors cur-
rently being deployed in the North Sea. After joining
the company in 1974, he spent four years as a eld
engineer in the Middle East. Since then he has
worked in various eld management, marketing and
interpretation support positions in Scotland,
Denmark and Norway, where he also worked as a
consultant for the Norwegian Petroleum Directorate.
During the last 15 years he has developed an extensive
knowledge base of formation evaluation methods
using the latest logging-while-drilling and wireline
sensors, particularly when applied to the reservoirs
of northwest Europe. Jim has a BS degree in physics
from Imperial College of Science, Technology and
Medicine in London.
Dean Willberg is Senior Program Manager and
Principal Engineer for the Schlumberger Well Product
Services Center in Moscow and at the Novosibirsk
Technology Center, where he works on developing
hydraulic fracturing and stimulation technologies.
He joined the company at the Tulsa Technology
Center in 1996 as a product development engineer
and later worked in product development at the Sugar
Land Product Center, Texas. Dean holds a BS degree
from the University of Alberta, Edmonton, and a PhD
degree from California Institute of Technology,
Pasadena, both in chemistry.
Wei Zhou, Schlumberger Well Services Sales Manager
for the China, Japan, Korea, Taiwan (CHG)
GeoMarket region, is based in Beijing. Before this, he
was an in-house DESC engineer for BP in Stavanger,
where he provided technical support to CT SEAS*
Coiled Tubing Safer, Efcient Automated Solutions
and to conventional coiled tubing units for the BP
Valhall eld operation. He introduced and engineered
the PowerCLEAN cleanout system to the Valhall ank
development post-fracture proppant cleanout process.
Wei joined Dowell Schlumberger in 1997 in Shekou,
China, as a coiled tubing eld engineer and later
worked in Prudhoe Bay, Alaska, as a general eld
engineer before moving to Stavanger. He is a graduate
of Tsinghua University, Beijing, with a BS degree
in engineering.
An asterisk (*) is used to denote a mark of Schlumberger.
Casing Drilling is a registered trademark of Tesco Corporation.
Oileld Review
NEW BOOKS
C ritical Mass: How One Thing
Leads to Another
Philip Ball
Farrar, Straus and Giroux
19 Union Square West
New York, New York 10003 USA
2004. 520 pages. $27.00
ISBN 0-374-28125-4
Covering a wide variety of topics, this
book outlines the value of scientific
study in increasing the understanding
of human behavior. Investigating
innovative attempts to explain social
behavior by applying formulas borrowed
from physics, the author explains how
some social theorists are now using
discoveries about molecular motion and
crystal formation to predict the
behavior of various human groups,
including crowds of soccer fans and
clusters of pedestrians.
Contents:
Introduction: Political Arithmetick
Raising Leviathan: The Brutish
World of Thomas Hobbes
Lesser Forces: The Mechanical
Philosophy of Matter
The Law of Large Numbers:
Regularities from Randomness
The Grand Ah-Whoom: Why Some
Things Happen All at Once
On Growth and Form: The
Emergence of Shape and
Organization
The March of Reason: Chance and
Necessity in Collective Motion
On the Road: The Inexorable
Dynamics of Traffic
Rhythms of the Marketplace: The
Shaky Hidden Hand of Economics
Agents of Fortune: Why Interaction
Matters to the Economy
Uncommon Proportions: Critical
States and the Power of the
Straight Line
The Work of Many Hands: The
Growth of Firms
Join the Club: Alliances in
Business and Politics
Multitudes in the Valley of
Decision: Collective Influence and
Social Change
Summer 2005
The Colonization of Culture:
Globalization, Diversity, and
Synthetic Societies
Small Worlds: Networks That Bring
Us Together
Weaving the Web: The Shape of
Cyberspace
Order in Eden: Learning to
Cooperate
Pavlovs Victory: Is Reciprocity
Good for Us?
Toward Utopia: Heaven, Hell, and
Social Planning
Epilogue: Curtain Call
Notes, Bibliography, Index
Some 17 pages of notes, a
13-page bibliography, and a 16-page
index round out this fascinating book,
appealing to both scientists and
humanists at every levelHighly
recommended.
Howard WE III: Choice 42, no. 4
(December 2004): 681.
Modern Problems in Classical
Electrodynamics
Charles A. Brau
Oxford University Press
198 Madison Avenue
New York, New York 10016 USA
2004. 594 pages. $99.95
ISBN 0-19-514665-4
Designed as an upper-level
undergraduate/beginning graduate
textbook and as a reference for
research scientists, this book addresses
a wide range of topics in modern
physicselectrostatics, magneto-
statics, waves, continuous media,
nonlinear optics, diffraction, radiation
by moving particles and the classical
theory of electrons. It features examples
and problem exercises drawn from
condensed-matter physics, particle
physics, optics and atomic physics.
Contents:
Prologue
Relativistic Kinematics
Relativistic Mechanics and
Field Theory
Time-Independent
Electromagnetic Fields
Electromagnetic Waves
Fourier Techniques and
Virtual Quanta
Macroscopic Materials
Linear, Dispersive Media
Nonlinear Optics
Diffraction
Radiation by Relativistic Particles
Fundamental Particles in Classical
Electrodynamics
Appendix, Index
The main contribution of this
bookis the inclusion of sections on
nonlinear optics and lasers, topics
that are rarely covered in a classical
electrodynamics book. The book also
provides a number of problems that
should prove valuable both to the
studentand to the instructor.
It should also be useful
for the student who wants to go
beyond the regular graduate
electrodynamics course.
Javier Gonzalez H: The Industrial Physicist 10,
no. 3 (JuneJuly 2004): 3334.
Detachment Fault
Susan Cummins Miller
Texas Tech University Press
Box 41037
Lubbock, Texas 79409 USA
2004. 235 pages. $24.95
ISBN 0-89672-520-0
The author, a former field geologist with
the US Geological Survey, also taught
geology and oceanography before
becoming a writer of fiction and nonfic-
tion. Detachment Fault is her second
murder mystery featuring Frankie
MacFarlane, a Tucson-based geologist
who becomes involved in solving a
murder. Like her first book, Death
Assemblage, this work includes
descriptions of Arizona natural history,
landscapes and geology.
In another geology murder
mysterythe second by geologist
Susan Cummins MillerTucson
geologist Frankie MacFarlane
investigates her second murder.
These are fun and easy reads,
interspersed with some science.
Geotimes 50, no. 1 (January 2005): 47.
Earth: An Intimate History
Richard Fortey
Alfred A. Knopf, a division
of Random House
201 E. 50th Street
New York, New York 10022 USA
2004. 429 pages. $30.00
ISBN 0-375-40626-3
The Alps, the Scottish Highlands,
Newfoundland and the Deccan Traps of
India are among the authors destina-
tions as he explains the theory of plate
tectonics to account for the formation
of the continents and oceans. Richard
Fortey, a senior paleontologist at
Londons Natural History Museum, has
written a series of related essays that
provide an insiders view of our planet
and its geologic history.
Contents:
Up and Down
Island
Oceans and Continents
Alps
Plates
Ancient Ranges
The Dollar
Hot Rocks
Fault Lines
The Ancient of Days
Cover Story
Deep Things
World View
Further Reading, Index
This is perhaps the most valuable
aspect of the book for the professional
reader: to have the historical sequence
of the workers who developed the fun-
damental tenets of mountain building,
geologic time, and the like laid out is a
real treat. Fortey has written a truly
enjoyable book for the professional
and lay reader alikeit should be
welcome reading for anyone with an
interest in the natural sciences.
Strother PK: Choice 42, no. 8 (April 2005): 1429.
65
Coming in Oilfield Review
Managing Gas-Condensate
Reservoirs. A retrograde gas-
condensate fluid drops out liquid
hydrocarbon when the fluid drops
below its dewpoint pressure. The
condensation can occur in the
formation and create a condensate
bank that decreases production; or
it can happen in the wellbore, loading
up the well with the heavier phase
and often requiring intervention to
maintain production. This article
describes efforts to maintain well
productivity despite declining
reservoir pressure.
The Pressures of Drilling and
Production. Geophysical pressure
development is rooted in the Earths
beginning. Millions of years later,
E&P companies predict, measure
and manage pressure while drilling
wells and producing reservoirs. This
article reviews the development of
geopressure systems, and then
discusses the risk and interdepen-
dencies of formation pressure on
drilling, production and hydrocarbon
recovery. Case histories show how
drillers and engineers are using
advanced techniques for pressure
prediction, detection and management
allowing wells to be drilled more
safely, boreholes to be placed more
accurately, and reservoirs to be
managed for maximum oil and
gas recovery.
Formation Evaluation While
Drilling. A higher standard for
formation evaluation while drilling
has now been established. An
innovative new tool design makes
thorough formation evaluation
possible without the use of chemical
radioactive sources, reduces the rig
time used to make up and break
down the bottomhole assembly, and
enables higher rates of penetration
during logging. Highlighted in this
article, the integrated logging-while-
drilling collar delivers new measure-
ments, along with significant safety
and efficiency advantages, to asset
teams around the world.
66
Pandoras Breeches: Women,
Science and Power in the
Enlightenment
Patricia Fara
Pimlico
Random House
20 Vauxhall Bridge Road
London SW1V 2SA England
2004. 224 pages. $20.00 (paperback)
ISBN 1-8441-3082-7
Written by a historian of science at
Cambridge University, the book traces
womens participation in European
science during the 17th and 18th
centuries. The author has selected
eight women whose lives were entwined
with those of famous male scientists
during the Enlightenment; she also
interweaves stories of many other
women who also made significant
contributions to the sciences.
Contents:
Pandora/Eve/Minerva: Prologue;
Women/Science; Lady
Philosophy/Francis Bacon
In the Shadows of Giants: Elisabeth
of Bohemia/Ren Descartes;
Anne Conway/Gottfried Leibniz;
milie du Chtelet/Isaac Newton
Domestic Science: Jane Dee/John
Dee; Elisabetha Hevelius/Johannes
Hevelius; Caroline Herschel/
William Herschel; Marie Paulze
Lavoisier/Antoine Lavoisier
Under Sciences Banner: Priscilla
Wakefield/Carl Linnaeus; Mary
Shelley/Victor Frankenstein; Epilogue
Notes, Bibliography, Index
Undermining the concept of
heroism in science, Pandoras Breeches
represents subaltern history of science
at its best. The stories Fara has
collected should also encourage the
recovery and reappraisal of the
contributions of many other scientific
workers who are now forgotten.
Gopinathan A: Science 307, no. 5709
(January 28, 2005): 522.
Understanding Renewable
Energy Systems
Volker Quaschning
Earthscan Publications Ltd.
8-12 Camden High Street
London NW1 0JH England
2005. 272 pages. $39.95 paperback;
$155.00 hardcover
ISBN 1-8440-7128-6
This book provides basic information
about various renewable energy systems,
including their function and practicality
in various applications. This updated
English version of the classic 1998 text
includes a CD-ROM with computer sim-
ulations that support the text. The top-
ics include a basic rationale for using
renewable energy (such as biomass,
hydroelectricity, geothermal and tidal
systems), and some of the most promis-
ing systems such as solar radiation,
photovoltaics and wind power. The
economics of renewable energy
resources is also discussed.
Contents:
Energy, Climate Change and
Renewable Energy Sources
Solar Radiation
Solar Thermal Water Heating
Photovoltaics
Wind Power
Economics
Simulation Programs and the CD-
ROM of the Book
Appendix, Bibliography, Index
Excellent illustrations and
graphs are complemented by an
accompanying CD-ROM that displays
all of the figures in the book plus 19
relevant software programs. The
bibliography is extensive. This work
is for all students of the worlds
energy dilemma.
Comer JC: Choice 42, no. 9 (May 2005): 162.
The Big One: The Earthquake
That Rocked Early America and
Helped Create a Science
Jake Page and Charles Officer
Houghton Mifflin Company
215 Park Avenue South
New York, New York 10003 USA
2004. 239 pages. $24.00
ISBN 0-618-34150-1
The three largest earthquakes to strike
the continental United States occurred
in 1811 and 1812 near New Madrid,
Missouri. This book provides a brief his-
tory of the science of seismology and a
basic primer on the current state of
geological knowledge. Using eyewitness
accounts and new scientific findings,
the authors tell the story of the quakes
and present theories on their causes,
outlining how the New Madrid earth-
quakes contributed to the creation of
the modern science of seismology.
Contents:
The New Madrid Quakes: The World
Gone Mad; Dreams, Omens, and
War; Pendulums and Polymaths
The Earthquake Hunters: Myths,
Maps, and Machines; Finding
Faults; Intensity, Magnitude, and
Stars; Geophysical Leaps Forward
Looking Back, Looking Forward:
Rifts, Plumes, and Reservoirs; The
Art of Prediction; False Prophets;
New Madrid Redux
Notes, Index
The book is written to be enjoyed
by both earth scientists and those not
in this field, and in this respect the
book is completely successful. There
are excellent explanations of the
development of seismic methods for
earthquake study. References are given
for technical and non-technical
publications.
Haberfield J: AAPG Bulletin 89, no. 4
(April 2005): 551552.
Oilfield Review